Train Crew Staffing

DEPARTMENT OF TRANSPORTATION

Federal Railroad Administration

49 CFR Part 218

[Docket No. FRA-2014-0033, Notice No. 1]

RIN 2130-AC48

Train Crew Staffing

AGENCY:

Federal Railroad Administration (FRA), Department of Transportation (DOT).

ACTION:

Notice of proposed rulemaking (NPRM).

SUMMARY:

FRA proposes regulations establishing minimum requirements for the size of train crew staffs depending on the type of operation. A minimum requirement of two crewmembers is proposed for all railroad operations, with exceptions proposed for those operations that FRA believes do not pose significant safety risks to railroad employees, the general public, and the environment by using fewer than two-person crews. This proposed rule would also establish minimum requirements for the roles and responsibilities of the second train crewmember on a moving train, and promote safe and effective teamwork. Additionally, FRA co-proposes two different options for situations where a railroad wants to continue an existing operation with a one-person train crew or start up an operation with less than two crewmembers. Under both co-proposal options, a railroad that wants to continue an existing operation or start a new operation with less than a two-person train crew would be required to describe the operation and provide safety-related information to FRA; however, proposed Option 1 includes an FRA review and approval period lasting up to 90 days while Option 2 proposes permitting such operations to initiate or continue without a mandatory FRA review and approval waiting period or while such review is taking place. For start-up freight operations with less than two crewmembers, proposed Option 2 also requires a statement signed by the railroad officer in charge of the operation certifying a safety hazard analysis of the operation has been completed and that the operation provides an appropriate level of safety.

DATES:

(1) Written Comments: Written comments on the proposed rule must be received by May 16, 2016. Comments received after that date will be considered to the extent possible without incurring additional expense or delay.

(2) FRA anticipates being able to resolve this rulemaking without a public, oral hearing. However, if FRA receives a specific request for a public, oral hearing prior to April 14, 2016, one will be scheduled and FRA will publish a supplemental notice in the

Federal Register

to inform interested parties of the date, time, and location of any such hearing.

ADDRESSES:

You may submit comments identified by the docket number FRA-2014-0033 by any of the following methods:

•

Online:

Comments should be filed at the Federal eRulemaking Portal,

http://www.regulations.gov.

Follow the online instructions for submitting comments.

•

Fax:

202-493-2251.

•

Mail:

Docket Management Facility, U.S. Department of Transportation, 1200 New Jersey Avenue SE., W12-140, Washington, DC 20590.

•

Hand Delivery:

Room W12-140 on the Ground level of the West Building, 1200 New Jersey Avenue SE., Washington, DC 20590 between 9 a.m. and 5 p.m., Monday through Friday, except Federal Holidays.

Instructions:

All submissions must include the agency name, docket name and docket number or Regulatory Identification Number (RIN) for this rulemaking (RIN 2130-AC48). Note that all comments received will be posted without change to

http://www.regulations.gov,

including any personal information provided. Please see the Privacy Act heading in the

Supplementary Information

section of this document for Privacy Act information related to any submitted petitions or materials.

Docket:

For access to the docket to read background documents or comments received, go to

http://www.regulations.gov

at any time or to the U.S. Department of Transportation, Docket Operations, M-30, West Building, Ground Floor, Room W12-140, 1200 New Jersey Avenue SE., Washington, DC 20590, between 9 a.m. and 5 p.m., Monday through Friday, except Federal Holidays.

FOR FURTHER INFORMATION CONTACT:

Joseph D. Riley, Railroad Safety Specialist (OP)-Operating Crew Certification, U.S. Department of Transportation, Federal Railroad Administration, Mail Stop-25, Room W33-412, 1200 New Jersey Avenue SE., Washington, DC 20590, (202) 493-6318, or Alan H. Nagler, Senior Trial Attorney, U.S. Department of Transportation, Federal Railroad Administration, Office of Chief Counsel, RCC-10, Mail Stop 10, West Building 3rd Floor, Room W31-309, 1200 New Jersey Avenue SE., Washington, DC 20590, (202) 493-6038).

SUPPLEMENTARY INFORMATION:

Table of Contents for Supplementary Information

I. Executive Summary

II. Background

A. Analysis of Two Recent Catastrophic Accidents Raising Crew Size Issues

1. Lac-Mégantic, Quebec, Canada

2. Casselton, ND

B. Research Identifies Crewmember Tasks and the Positive Attributes of Teamwork, Raises Concerns With One-Person Crews, Especially When Implementing New Technology

1. Cognitive and Collaborative Demands of Freight Conductor Activities: Results and Implications of a Cognitive Task Analysis—Human Factors in Railroad Operations

2. Rail Industry Job Analysis: Passenger Conductor

3. Fatigue Status in the U.S. Railroad Industry

4. Technology Implications of a Cognitive Task Analysis for Locomotive Engineers—Human Factors in Railroad Operations

5. Using Cognitive Task Analysis To Inform Issues in Human Systems Integration in Railroad Operations—Human Factors in Railroad Operations

6. Teamwork in U.S. Railroad Operations

C. The Acknowledged Limitations of FRA Accident/Incident Reporting Data

D. FRA's Regulations Suggest Safety Hazards Are Created When a Train Has Less Than Two Crewmembers

1. Difficulty Providing Point Protection for Shoving or Pushing Movements

2. Complications Returning Switches to the Normal Position and Loss of Job Briefings

3. Concerns Protecting Train Passengers in an Emergency

4. Deterrence of Electronic Device Distraction and Observing Alcohol or Drug Impairment, Reduced Possibility of Co-Worker Referrals

5. Complicating Radio Communication Procedures

6. Adding a Potential Safety Hazard to Highway-Rail Grade Crossing Activation Failures

E. Defining the Crewmembers' Qualifications

III. Railroad Safety Advisory Committee (RSAC) Overview

IV. No Recommendation From the RSAC Working Group

V. FRA's Overall Post-RSAC Approach

A. The Proposal Is Largely Focused on Influencing How Railroads Approach Future One-Person Operations

B. The Proposal Is Complimentary to Other Regulatory Initiatives, Not Duplicative

C. Identifying How the NPRM Differs From FRA's RSAC Suggested Recommendations

D. Electronic Submission and Approval Process

VI. Section-by-Section Analysis

VII. Regulatory Impact and Notices

A. Executive Order 12866, Executive Order 13563, and DOT Regulatory Policies and Procedures

B. Regulatory Flexibility Act and Executive Order 13272

C. Paperwork Reduction Act

D. Federalism Implications

E. International Trade Impact Assessment

F. Environmental Impact

G. Unfunded Mandates Reform Act of 1995

H. Energy Impact

I. Privacy Act

I. Executive Summary

Purpose of the Regulatory Action and Legal Authority

FRA is concerned that as railroads implement positive train control (PTC) and other technologies, they may expand use of less than two-person crews on operations without considering safety risks or implementing risk mitigating actions that FRA believes are necessary. Because there are currently few railroad operations that utilize a one-person crew and FRA has not been specifically tracking the safety of those operations through its recordkeeping and reporting requirements, FRA cannot provide reliable or conclusive statistical data to suggest whether one-person crew operations are generally safer or less safe than multiple-person crew operations. FRA does not currently collect sufficient data related to the size of a train crew nor do accident reports and investigations generally address the size of a crew in order for FRA or any entity to definitively compare one-person operations to multiple person operations. However, FRA has studies showing the benefits of a second crewmember and other information detailing the potential safety benefits of multiple-person crews. A recent catastrophic accident in Canada occurred in which a one-person crew did not properly secure an unattended train and another accident occurred in which a multiple-person crew was able to effectively respond to an accident and remove cars from danger. In addition, qualitative studies show that one-person train operations pose increased risks by potentially overloading the sole crewmember with tasks, and that PTC does not substitute for all the tasks performed by properly trained conductors. Task overload can lead to a loss of situational awareness, and potentially to accidents. Moreover, other nations require government approval of railroad decisions to use less than two-person crews. Further, even if FRA does not have data to prove a direct correlation between higher rates of safety and multiple person crews, it is true that railroads have achieved a continually improving safety record during a period in which the industry largely employed two-person train crews.

Persons in the railroad industry have pointed to countervailing effects of a requirement to have more than one crewmember on a train, such as additional incidents caused by crew distraction. In addition, having a second crew person on board a train may not prevent or mitigate an incident but could add to the number of persons killed or seriously injured when one occurs. FRA believes such instances are very rare, but does not have readily available information for estimating such potential countervailing impacts of this proposed rule. FRA believes that having a properly trained second crew person on board, or implementing risk mitigating actions that FRA believes are necessary to address any additional safety risks from using fewer than two-person crews, provides net safety benefits relative to using fewer than two-person crews or not implementing mitigating measures that FRA believes are necessary.

In discussing the future of train operations with officials from various railroads, FRA has become aware that some railroads have shown a willingness to conduct more operations with only one crewmember. FRA has existing authority to take emergency action to prohibit an unsafe operation if the agency is aware of it (49 U.S.C. 20104), but FRA often lacks information to use this authority to address unsafe one-person crews. FRA does not currently have a mechanism to collect detailed information about railroad one-person train operations to determine railroad safety risk. Furthermore, FRA believes it would be inappropriate to wait until an emergency situation arises before it takes action against a one-person operation that is not providing an appropriate level of safety. FRA believes this proposed rule is necessary for FRA to protect railroad employees, the general public, and the environment by considering the safety risks of each type of operation and prohibiting operations that pose an unacceptable level of risk as compared to operations utilizing a two-person crew. This rulemaking is also necessary to ensure that the public, through FRA, has a voice in the railroad's decision to utilize less than a two-person crew.

FRA research demonstrates the effectiveness of properly trained teams. It is not the act of adding a second person that makes the train safer, but instead it is the act of adding a properly qualified person, who understands the roles of all the crewmembers, and who has the experience or ability to relieve the locomotive engineer of some of the mental strain that can contribute to accidents attributed to human factor errors. FRA understands that expert teamwork can be achieved through effective coordination, cooperation, and communication. However, FRA estimates both options of the proposal would have a small impact on teamwork because FRA expects that either co-proposal option would result in no more than the labor hour equivalent of two to three additional crewmembers nationwide annually relative to what would occur with existing operations with less than two crewmembers if the rule were not in place and because FRA believes that all railroads with multiple-person crews are operating in compliance with the proposal's requirements for the roles and responsibilities of a second crewmember. FRA expects that under the first co-proposal it would require some start-up one-person crew operations (but not existing one-person crew operations) to implement risk mitigating measures that FRA believes are necessary to address safety risks of using one-person crews in specific operating environments. However, FRA expects to require such measures in very few circumstances, and estimates a cost range of $5.1 million to $27.7 million over 10 years and discounted at 7 percent from implementing such measures under either co-proposal option.

The proposed rulemaking would be expected to grant an exception to most existing operations with less than two crewmembers. However, some operations would still not be able to meet the requirements of the proposed exceptions and those railroads would have to add one person to their train crews. FRA estimates that about 10,361 train starts would not be eligible for the proposed specific freight train exception § 218.131. Furthermore, FRA estimated that around 15,185 train starts would not be covered by the exception for existing one-person operations in § 218.133. Given the proposed structure of the passenger train exceptions in § 218.129, FRA does not expect any passenger railroad to have to add a crewmember to an existing train operation as a result of the NPRM. Freight railroads would be expected to take full advantage of the special approval procedure in § 218.135. FRA used a range of values to estimate the costs that would be related to § 218.135 due to the uncertainty in the future of crew staffing. This range stipulates that

between 850,266 and 15,675,000 train starts would be affected by crew reduction over the next 10 years and enter the special approval procedure as proposed in § 218.135. For passenger railroads, the proposed special approval procedure would maintain the status quo, as any railroad that could potentially request special approval under § 218.135 would have done it through a passenger train emergency preparedness plan under part 239.

FRA is proposing regulations concerning train crew staffing based on the statutory general authority of the Secretary of Transportation (Secretary). The general authority states, in relevant part, that the Secretary “as necessary, shall prescribe regulations and issue orders for every area of railroad safety supplementing laws and regulations in effect on October 16, 1970.” 49 U.S.C. 20103. The Secretary delegated this authority to the Federal Railroad Administrator. 49 CFR 1.89(a).

Summary of the Major Provisions of the Regulatory Action in Question

FRA is co-proposing regulations to address train crew sizes. FRA's first co-proposal would establish minimum requirements for the size of different train crew staffs depending on the type of operation and the safety risks posed by the operation to railroad employees and the general public. This proposal also prescribes minimum requirements for the appropriate roles and responsibilities of train crewmembers on a moving train, and promotes safe and effective teamwork. Each railroad may prescribe additional or more stringent requirements in its operating rules, timetables, timetable special instructions, and other instructions.

FRA's first proposed approach starts with a general requirement that each train shall be assigned a minimum of two crewmembers, regardless of whether the train is a freight or passenger operation. The NPRM contains several proposed requirements detailing the roles and responsibilities of the second crewmember when the train is moving. The primary role of a second crewmember, typically a conductor, is to have the ability to directly communicate with the crewmember in the cab of the controlling locomotive,

i.e.,

the locomotive engineer, even if the second crewmember is located outside of the operating cab.

Several of the proposed sections contain exceptions to this general requirement, specifying when a train would not require a minimum of two crewmembers. These are generally low risk operations that are not hauling large quantities of hazardous materials, traveling at high speeds, or putting passengers on passenger trains at risk. Among other exceptions, there is a proposed exception for a tourist, scenic, historic, or excursion operation that is not part of the general railroad system of transportation. Other exceptions allow railroads to use one-person crews to assist other trains (

i.e.,

helper service), maintain track, or move locomotives where they are needed without being burdened by the proposed two crewmember minimum staffing requirement.

Two of the proposed sections suggest how a railroad could apply for FRA approval to operate one-person train crews. One of those proposed sections would require a railroad to provide information describing an operation that existed prior to January 1, 2015, and FRA would have 90 days from the day of receipt of the submission to issue written notification of approval or disapproval. The railroad would be allowed to continue the operation unless FRA notifies the railroad it must cease the operation and provides the reason(s) for the decision. If FRA failed to disapprove the proposal within 90 days of the submission, the railroad would be permitted to go forward with its plan. The second of the proposed sections under the first co-proposal would allow any railroad, at any time, to provide information describing an operation and petition FRA for special approval of a train operation with less than two crewmembers. FRA would normally grant or deny the petition within 90 days of receipt, but could attach special conditions to the approval of any petition after considering the benefits and costs of the condition(s).

Under the second co-proposal, an existing one-person train operation would be required to provide information to FRA in order to continue the operation, and a start-up train operation with less than two crewmembers would be required to provide information to FRA before initiating the operation. The railroad with the start-up operation would also be required to attest that it has studied the operating environment and circumstances of the intended operation and that the railroad believes that it has taken any precautions necessary to ensure that the proposed single-person operation will not pose significant safety risks to railroad employees, the general public, and the environment. Under this co-proposal, the railroad would not be required to wait for FRA approval prior to beginning single-person service. With the railroad's notice and attestation the railroad would be permitted to operate a single-person service. Both existing and start-up train operations with less than two crewmembers would be required to provide an appropriate level of safety. However, FRA reserves the right to investigate an operation and halt or add conditions to an operation's continuance if FRA determines that an operation is not providing an appropriate level of safety.

Costs and Benefits

FRA estimated the benefit and cost ranges of the two co-proposals using a 10-year time horizon, and performed sensitivity analysis using a 20-year time horizon. Compliance costs include the addition of the labor hour equivalent of about one to three additional crewmembers nationwide annually to certain train movements for existing operations (an estimated cost of roughly $120,000-$200,000 annually over 10 years, undiscounted), off-setting actions required by FRA in order for a railroad to obtain FRA approval to start up new fewer than two-person crew operations, and information submission and data analysis.

FRA estimated a 10-year cost range which would be between $7.65 million and $40.86 million, undiscounted. Discounted values of this range are $5.19 million and $27.72 million at the 7-percent level. FRA is confident that the benefits outlined in this NPRM would exceed the costs. Preventing a single fatal injury would exceed the break-even point in the low range and preventing five fatalities would exceed the break-even point at the high range. The proposed rule will help ensure that train crew staffing does not result in inappropriate levels of safety risks to railroad employees, the general public, and the environment, while allowing technology innovations to advance industry efficiency and effectiveness without compromising safety. The proposal contains minimum requirements for roles and responsibilities of second train crewmembers on certain operations and promotes safe and effective teamwork. Due to lack of information, these cost estimates do not include any safety costs from using two-person crews instead of one or zero person crews, such as additional accidents caused by non-engineer crew distracting the engineer or additional deaths and serious injuries from having more people on board trains involved in accidents.

FRA is confident that the proposed rulemaking would generate the benefits necessary to at least break-even. These benefits would result from improved

post-accident/incident emergency response and management, reporting of troubled employees due to drug and alcohol use, compliance with restrictions on electronic device use in place to prevent distraction, and the potential avoidance of a high-consequence train accident. While FRA does not have information that suggests that there have been any previous accidents involving one-person crew operations that could have been avoided by adding a second crewmember, this rule would break even with its estimated costs if it prevents one fatal injury or high-consequence accident in the first 10 years of the rule (and no additional safety costs result from the presence of additional crew). This proposed rule would help ensure that passengers and high risk commodities are transported safely by rail and FRA is confident that the resulting safety benefits would justify the costs. The cost increase would result from additional crewmembers on the trains that are currently operating with a one-person crew and from the possibility that the railroad is required to use more technology to mitigate the risk related to crew conversions. FRA has assessed both co-proposals and concluded that monetary, quantifiable costs under both co-proposals are equal. However, railroads may perceive each option differently, especially as it pertains to business risk. Under co-proposal Option 1, railroads would have to wait for approval and that would delay implementation of crew size reduction in the short-term. However, once FRA grants approval railroads would have spent adequate amount of resources to meet regulatory requirements and oversight. Under co-proposal Option 2, each railroad would be able to initiate crew reductions after a petition is submitted to FRA. This means that railroads would be able to reduce costs once petitions are submitted. However, under co-proposal Option 2, railroads may assume more business risk as an initiated crew reduction would be subject to regulatory action (discontinuance or more conditions for approval). This means that railroads could end up acquiring equipment or resources for unapproved crew reductions or to modify initial plans for crew reductions. This would be costly and bring more uncertainty to the railroads' business plans in the short-term.

FRA conducted a sensitivity analysis of its first co-proposal using a 20-year time horizon and a scenario with a more rapid crew size reduction schedule. FRA estimates that the cost range of the co-proposals would be $7.44 million to $36.25 million over this timeframe using a 7-percent discount rate, and $11.93 million to $50.71 million using a 3-percent discount rate.

II. Background

A. Analysis of Two Recent Catastrophic Accidents Raising Crew Size Issues

During the last five months of 2013, the railroad industry had two accidents that suggest the need for greater Federal oversight of crew size issues. The first incident at Lac-Mégantic, Quebec, Canada, was the driving force for bringing the crew size issue to FRA's Federal advisory committee known as the Railroad Safety Advisory Committee (RSAC). While Canada's Transportation Safety Board could not conclude that use of a one-person crew was a cause or contributing factor to the accident, as described below, the Lac-Mégantic accident involved a one-person crew that did not properly secure a train at the end of a tour of duty leading to a deadly, catastrophic accident.

The RSAC includes representatives from all of the agency's major stakeholder groups, including railroads, labor organizations, suppliers and manufacturers, and other interested parties. (An RSAC overview is provided below.) During the time that the RSAC's Working Group was deliberating whether it could make recommendations to FRA on the crew size issue, the other accident summarized here occurred. This accident involved trains carrying multi-person crews and is illustrative of the positive mitigation measures multiperson train crews took following a track-based derailment of one train that led to a second train colliding with the first (Casselton, ND). With regard to the Lac-Mégantic accident, FRA exercised its oversight following the accident through use of its emergency order authority to ensure that the railroad involved had at least one adequate backstop to human error. FRA has also issued several other regulations to address the safety issues raised by these accidents which are described within the summaries of the accidents.

Lac-Mégantic, Quebec, Canada

FRA published Emergency Order 28 (78 FR 48218) on August 7, 2013, (issued on August 2, 2013) which contains the preliminarily known details of the events on July 5-6, 2013, that led to the catastrophic accident at Lac-Mégantic. On August 20, 2014, the Transportation Safety Board (TSB) of Canada released its railway investigation report, which refines the known factual findings and makes recommendations for preventing similar accidents. TSB of Canada Railway Investigation R13D0054 is available online at

http://bit.ly/VLqVBk.

In summary, an unattended train on mainline track did not stay secured and rolled down a grade to the center of town, where 63 of the 72 crude oil tank cars in the train derailed, and about one-third of the derailed tank car shells had large breaches. There were multiple explosions and fires causing an estimated 47 fatalities to the general public, extensive damage to the town, and approximately 2,000 people to be evacuated from the surrounding area.

The train had been secured by its one-person crew prior to it being left unattended. Because of a mechanical problem with the train, the engineer left the train running. Prior to leaving the train, the engineer consulted with another railroad employee about how to handle the problem and applied brakes on the train. However, TSB of Canada determined that the one-person crew did not comply with the railroad's rules requiring the hand brakes alone to be capable of holding the train. According to the railroad's rules, a 72-car train should have had a minimum of nine hand brakes applied. Instead, the one-person crew used a combination of the locomotive air brakes and seven hand brakes to give the false impression during the verification test that the hand brakes alone would hold the train. TSB of Canada concluded that, without the extra force provided by the air brakes, a minimum of 17 and possibly as many as 26 hand brakes would have been needed to secure the train, depending on the amount of force with which they had been applied. Testing conducted by TSB of Canada concluded that it would have been possible for a single operator to apply a sufficient number of hand brakes within a reasonable amount of time. Shortly after the one-person crew left the train, the local fire department responded to an emergency call about a fire on the train. The responders followed the railroad's instructions in shutting down the locomotive and then extinguished the fire. The responders met with an employee of the railroad, a track foreman, to discuss the train's condition prior to departing the area. The track foreman dispatched by the railroad did not have a locomotive operations background. With all the locomotives shut down, the air compressor no longer supplied air to the air brake system, the air leaked, and the air brakes gradually become less effective until the combination of

locomotive air brakes and hand brakes could no longer hold the train.

In the aftermath of the Montreal, Maine and Atlantic Railway (MMA) derailment at Lac-Mégantic, Transport Canada issued an order for all Canadian railroad companies to provide for minimum operating crew requirements considering technology, length of train, speeds, classification of dangerous goods being transported, and other risk factors. In response, MMA changed its operating procedures to use two-person crews on trains in Canada. However, FRA was concerned that MMA did not automatically make corresponding changes to its operating procedures in the U.S. even though the risk associated with this catastrophic accident also exists in the U.S.

1

It may have been that, without a specific two-person train crew requirement in the U.S., MMA did not feel compelled to take any action to enhance the safety of its U.S. operations in a like-minded way to the preventive measures it took in Canada.

1

Letter from Joseph C. Szabo, FRA Administrator, to Mr. Edward Burkhardt, CEO of MMA (Aug. 21, 2013), placed in the docket.

The Lac-Mégantic accident is also relevant to the issue of crew size because the tank cars that derailed were carrying crude oil from the Bakken deposit in North Dakota and Montana and this proposed rule carries forward FRA's position that at least a two-person train crew is warranted on any train carrying 20 or more tank cars loaded with crude oil or ethanol. Over the past few years, a technological advancement has allowed crude oil to be recovered from under nonpermeable shale rock. This advancement of hydraulic fracturing, better known as “fracking,” resulted in a substantial increase in crude oil shipments in both Canada and the U.S. between 2009 and 2015.

2

The prevalence of crude oil tank cars on U.S. railroads, and the volatility of some of the blended crude oil from different sources or mixed with the chemicals used in the fracking process, suggested that Bakken crude oil might have a significantly greater potential to be improperly classified and packaged for transportation. Investigators initially considered that improper classification and packaging was likely a contributing cause to the catastrophic result at Lac-Mégantic. Consequently, DOT has taken or is taking a variety of actions to address the issues created by transporting crude oil produced through fracking from various approaches.

See,

the following examples

2

https://www.eia.gov/dnav/pet/PET_MOVE_RAILNA_A_EPC0_RAIL_MBBL_M.htm.

• FRA's Emergency Order 28, 78 FR 48218, Aug. 7, 2013.

• FRA's Safety Advisory 2013-06, 78 FR 48224, Aug. 7, 2013, jointly issued with the Pipeline and Hazardous Materials Safety Administration (PHMSA) (discussing the circumstances surrounding the Lac-Mégantic accident and making certain safety-related recommendations to railroads and crude oil offerors).

• FRA's Safety Advisory 2013-07, 78 FR 69745, Nov. 20, 2013, jointly issued with PHMSA (reinforcing the importance of proper characterization, classification, and selection of a packing group for Class 3 materials and the corresponding requirements in the Federal hazardous materials regulations for safety and security planning after the Lac-Mégantic accident).

• FRA's Safety Advisory 2014-01, jointly issued with PHMSA, 79 FR 27370, May 13, 2014, (encouraging the use of railroad tank car designs with the highest level of integrity reasonably available).

• PHMSA's final rule, issued in coordination with FRA, “Hazardous Materials: Enhanced Tank Car Standards and Operational Controls for High Hazard Flammable Trains,” 80 FR 26643, May 8, 2015, (adopting new operational requirements for certain trains transporting large quantities of flammable liquids known as “high-hazard flammable trains” (HHFT), creating improvements in tank car standards, providing a sampling and classification program for unrefined petroleum-based products; and creating notification requirements).

• FRA's final rule “Securement of Unattended Equipment,” 80 FR 47349, Aug. 6, 2015, (adopting requirements to prevent unattended trains that carry crude oil, ethanol, poisonous by inhalation (PIH), toxic by inhalation (TIH), and other highly flammable contents from rolling away).

Also, in 2013, DOT launched Operation Safe Delivery (OSD), which is examining the entire system of crude oil delivery. OSD concluded, after months of unannounced inspections, testing, and analysis, that “the current classification applied to Bakken crude is accurate under the current classification system, but that the crude has a higher gas content, higher vapor pressure, lower flash point and boiling point and thus a higher degree of volatility than most other crudes in the U.S., which correlates to increased ignitability and flammability.”

See

OSD Update (July 23, 2014) summarizing PHMSA and FRA testing results of Bakken crude oil as of May 2014; available online at

http://1.usa.gov/1piQJB1.

Some people in the railroad industry view the accident at Lac-Mégantic as having nothing to do with crew size. They argue that there are potential safety benefits to single-person train operations, such as increased attentiveness by the lone operator because of the absence of a second crewmember on whom to rely. It is also said that there are fewer distractions from extraneous conversations. The TSB of Canada report on the Lac-Mégantic accident found that it could not be concluded that a one-person crew contributed to the accident. However, TSB of Canada found that the risk of implementing single-person train operations is a risk that must be addressed because it is related to unsafe acts, unsafe conditions, or safety issues with the potential to degrade rail safety. TSB of Canada concluded that addressing the risk of one-person operations is essential to preventing future similar accidents, even if the risk itself cannot be determined to directly have led to this accident.

Related to the risks associated with one-person operations, TSB of Canada found that MMA did not have a strong safety culture, which made MMA a poor candidate to implement one-person operations. For instance, TSB of Canada notes that an organization with a strong safety culture is generally proactive when it comes to addressing safety issues, and yet MMA was generally reactive. MMA had significant gaps between the company's operating instructions and how work was performed day-to-day. Furthermore, TSB of Canada's investigation found MMA had inadequate training, testing, and supervision. In contrast, an effective safety culture is characterized by an informed workforce where people understand the hazards and risks involved in their own operation and work continuously to identify and overcome threats to safety.

At the time of the accident, there were no rules or regulations preventing Canadian railroads from implementing one-person train operations. Thus, TSB of Canada concluded that the risks posed by one-person operations suggest that Transport Canada,

i.e.,

Canada's DOT, should consider whether each railroad has the measures in place to mitigate those risks by creating a process to approve and monitor each railroad's one-person operation plans. TSB of Canada reasoned that if one-person operations are implemented “without identifying all risks, and if mitigation measures are not implemented, an equivalent level of safety to that of multi-person crews will not be maintained.” Considering that

there are only two Canadian railroads that have operated using one-person operations, TSB of Canada seems to be making a prudent recommendation before one-person operations are more widely used throughout the Canadian rail system. This is the exact lesson learned that FRA would like to apply to U.S. rail operations through promulgation of this rulemaking.

Even though TSB of Canada was not able to conclude that having another crewmember would have prevented the accident, and certainly FRA agrees that this could not be determined with any absolute certainty, it is distinctly possible that a train crew with a minimum of two-persons would have had more options available to secure the train safely, thereby potentially posing less of a risk of a runaway train. This was an issue raised by some labor members of FRA's Federal advisory committee and has some support in TSB of Canada's report. For instance, a one-person crew was limited to where the train could be parked so that it would not block a grade crossing, where it is significantly more feasible operationally for a two-person crew to choose to split the train and park each part on a lesser grade than the choice left for the one-person crew. There are four main reasons why splitting a train is generally considered a two-person job: (1) If a one-person crew leaves the locomotive cab unoccupied and has not taken appropriate measures to secure the train, it could become a runaway; (2) even if the train is secure, some cars may move depending on the terrain, making it difficult for the one-person crew to go between cars at a desired location without applying hand brakes, which can be time-consuming and strenuous work; (3) depending on the length of the train, it could be time-consuming for the one crewmember to walk the train to get to the desired location for a cut and find that the car needs to move to release the coupler lock; and (4) when the one-person crew stops occupying the lead locomotive cab, the train and crew are more vulnerable to vandalism and malicious acts by trespassers who might actually want to operate the train. In addition, a second person might be needed to flag a grade crossing and it would be easier to reposition one or more cars with a second crewmember. Another issue that favors two-person crews is that a TSB of Canada survey determined that there were instances when MMA one-person crews applied less than the minimum number of hand brakes required by MMA's rules and that the minimum hand brake requirement was more consistently met when trains were operated by two crewmembers. This seems to be the case here, as the engineer only set seven hand brakes instead of the minimum of nine. Although TSB of Canada's investigation found that even nine hand brakes would not have been enough to hold the train, a second crewmember could have ensured proper securement if the railroad had issued proper instructions regarding the minimum number of hand brakes to apply. Even TSB of Canada's report summarizing its investigations of other shortline runaway train accidents that it investigated previously suggests that, without having another crewmember available, no other person had the opportunity to verify whether the train was properly secured. Additionally, although it is not unusual for some types of locomotives to smoke and that the engineer did contact the railroad and was told to leave the engine while it was smoking, TSB of Canada found that the taxi driver that questioned the decision to leave the locomotive in a smoking condition did not carry the same weight as a qualified railroad employee. Similarly, the one-person crew and the dispatcher did not discuss the MMA procedure requiring that a locomotive be shut down due to abnormal smoke, and TSB of Canada states that it is impossible to conclude whether the presence of another crewmember would have resulted in different actions to secure the train—although FRA believes it is impossible to exclude either.

Thus, in consideration of the safety concerns involved in the rail transportation of crude oil, the catastrophic accident at Lac-Mégantic serves as the trigger to create redundant safeguards that have a high potential of preventing other accidents. FRA's position is reinforced by research and review of accident information, which confirms that railroads that provide two qualified crewmembers, who can work as an effective team on those unit trains (which commonly consist of over 100 loaded tank cars of crude oil), improve the safety of those operations.

Casselton, ND

Another train accident illustrates how having multiple train crewmembers can improve safety for the general public and the crewmembers themselves. On December 30, 2013, an eastbound BNSF Railway (BNSF) “key train,” consisting of two head end locomotives, one rear distributive power unit (DPU), and two buffer cars on each end of 104 loaded crude oil cars, collided with a car from a westbound BNSF “grain train” that had derailed less than 2 minutes earlier from an adjacent main track. Thirteen cars in the middle of the 112-car grain train had derailed, most likely due to a broken axle on the 45th railcar, and that railcar ended up fouling the main track the key train was operating over. The collision derailed the key train's two leading locomotives, as well as the first 21 trailing cars behind the locomotives. After the collision, an estimated 474,936 gallons of crude oil was released from 18 loaded tank cars fueling a fire which caused subsequent explosions as the loaded oil tank cars burned. The local fire department had requested that nearby residents voluntarily evacuate immediately following the collision and approximately 1,500 residents did evacuate. The voluntary evacuation was lifted approximately 25 hours after the collision. There were no injuries to crewmembers, emergency responders, or the general public, but images and video of the burning railcars made the accident national news.

Many members of the general public who viewed the news accounts of burning wreckage may not be aware that the heroic actions of the grain train's crewmembers potentially prevented the environmental and property damages from being much worse, in addition to potentially shortening the evacuation period. The grain train was operated by a three-person crew, which included a locomotive engineer, a conductor, and a student locomotive engineer (

i.e.,

a conductor training to be a locomotive engineer). Post-accident, the grain train crew was approached by the Assistant Fire Chief of the Casselton Fire Department who asked whether the crew could assist the emergency responders by pulling a cut of tank cars away from the burning derailed cars. Upon receiving the request, a BNSF road foreman of engines consulted with the crew to see if the crewmembers believed it was safe to move the cars, which they did. The grain train's locomotive engineer and student locomotive engineer went to the DPU on the key train and the conductor and road foreman of engines went to the east to the nearest grade crossing and made a cut of an estimated 50 tank cars. The engineer and student engineer then pulled the cars about a quarter of a mile west away from the burning train.

Approximately 45 minutes after that move was completed, the Assistant Fire Chief met the grain train's crew again and asked if additional tank cars from the key train could be moved. The grain train's crew made contact with a BNSF trainmaster and communicated the request. The trainmaster told the crew that if the move could be completed safely, they had permission to proceed.

The student engineer borrowed the Assistant Fire Chief's fire protective clothing and walked within 10 car lengths of the fire and uncoupled approximately 20 additional cars from the burning train. Then, the locomotive engineer coupled to these cars and moved them to the west creating a safety gap of approximately 25 to 30 car-lengths from the burning cars.

Adding these two emergency response moves together, the grain train's crew was responsible for moving approximately 70 loaded crude oil cars in the key train out of harm's way. These urgent moves would have been much more time consuming and logistically difficult if the grain train was operated with only a one-person crew. For those reasons, there is a question of whether either of these emergency response moves would have been attempted with a one-person crew.

Meanwhile, it is arguable that the two-person key train crew benefited from each other's presence in the cab of the controlling locomotive. The crew helped each other through the emergency by issuing appropriate warnings and sharing tasks. First, the locomotive engineer was able to warn the conductor to get down and brace for impact 4 to 5 seconds before colliding with the derailed grain train railcar, and they both were able to get down on the floor and brace themselves. The conductor admitted that he had never been in a situation where a collision was imminent, and he did not know what he was supposed to do. Although a one-person crew would not need to warn another crewmember of an impending impact, this is an example of an expert crew working together. Second, after the impact, the crew was able to assess that they were not seriously injured, and it was the conductor who first noticed that their train was on fire when he looked out the window and was able to warn the locomotive engineer of that fact. This is a clear example of the benefit a second crewmember can provide. Without a second person, the engineer may not have realized that he was in immediate danger. Third, upon hearing this news, the engineer told the conductor to “grab your cell phone and run.” This is another example of effective teamwork during an emergency situation. Some people do not think as clearly as others during an emergency and, in this case, the engineer, with about 9 years of experience, recognized that it was important for him to instruct the conductor with less than 2 years of experience that the crew should have their cell phones to report information and to leave the locomotive quickly. Fourth, the engineer announced the collision by radio. Reporting the incident as quickly as possible is always crucial to getting first responders to the scene of an accident. By contacting the dispatcher on the railroad's radio, the engineer was taking an important precaution to ensure other railroad operations were not adversely impacted. Had this been a one-person crew, there is a question of whether the engineer might have desired to exit the locomotive first and then notify the dispatcher, assuming the engineer believed his life was in immediate danger. Having a second crewmember present working to exit the locomotive may have freed the engineer to report the accident. Fifth, the conductor attempted to exit the front door while the engineer was reporting the accident over the radio, but finding it jammed shut, the conductor departed the locomotive through the back door located behind the engineer's seat. The engineer soon followed the conductor as it was clearly determined to be the only viable way to exit the locomotive. As the crew escaped from the locomotive, the conductor described the heat from the fire as “intense.” The crew could not get away from the locomotive quickly as they found themselves in knee-deep snow immediately upon exiting the locomotive. About a minute after exiting the locomotive, it was engulfed in flames. Sixth, they ran together away from the train with the engineer using his cell phone on the run to call 911 and the conductor answering the dispatcher's call on the conductor's cell phone. Thus, the two crewmembers were able to simultaneously assist with providing different officials with information that would assist the railroad and first responders. Seventh, when the engineer found out local citizens were at the crash site, he strongly urged the local police to get those citizens away from the site because the oil train was just like the one in (Lac-Mégantic) Canada, and the deputy sheriff recognized the danger. These two crewmembers worked as a team in an emergency situation to divide up tasks, warn the dispatcher and local emergency responders, and protect each other's safety. Fortunately, neither crewmember suffered any serious injuries preventing them from escaping the damaged locomotive or running to safety. Certainly, with two crewmembers, there is the potential that both crewmembers could be hurt, but there is also the possibility that one crewmember could physically assist an injured colleague. FRA believes that, from a post-accident risk mitigation standpoint, this accident is illustrative of the safety benefits a second crewmember can provide and that railroad operations, railroad crewmembers, the environment, and the general public are better served by the availability of a second crewmember. As explained in relation to the Lac-Mégantic accident, it is often impractical to expect a one-person crew to split a train, and in the case of an accident, there are added concerns regarding a one-person crew's ability to maintain communications with the dispatcher and emergency personnel while performing this potentially dangerous emergency movement. For instance, although an employee is permitted to use a cell phone during emergency situations involving the operation of the railroad under 49 CFR 220.309(b), the employee would have to remember to grab it, and the dispatcher and emergency personnel might not know the employee's phone number. If the employee took a portable railroad radio while conducting the train splitting operation, there is a significant probability that the radio signal would not be strong enough to communicate with the dispatcher. These concerns also do not take into account the fact that FRA purposely prohibits the use of electronic devices during railroad operations as they can be distractions that lead to preventable injuries and accidents.

See

49 CFR part 220, subpart C. The benefits of a second crewmember following an accident may be especially useful when the commodities hauled pose significant risks, or a single crewmember is injured or is simply unable to perform as many tasks as quickly as two crewmembers.

B. Research Identifies Crewmember Tasks and the Positive Attributes of Teamwork; Raises Concerns With One-Person Crews, Especially When Implementing New Technology

Before FRA asked RSAC to consider accepting a crew size task, FRA was aware that some research revealed significant safety concerns with one-person crew operations. To aid the Working Group in its development of recommendations for appropriate crew size minimum standards, FRA provided five FRA-sponsored research reports, as well as one Transportation Research Board (TRB) conference report that contains presentations from multiple research reports, prior to the first meeting. This background offers a summary of the important findings of these reports, as well as a list of those reports presented, with an internet link to each report.

(1) “Cognitive and Collaborative Demands of Freight Conductor Activities: Results and Implications of a Cognitive Task Analysis—Human Factors in Railroad Operations,” Final Report, July 2012, DOT/FRA/ORD-12/13. The research and report was performed by the John A. Volpe National Transportation Systems Center. The report is available online at

http://www.fra.dot.gov/eLib/details/L04331.

A primary finding of this FRA-sponsored study is that conductors and locomotive engineers operate as a joint cognitive system. The findings indicate that the conductor and the locomotive engineer function as an integrated team that often operate as a single unit with a common goal. These two crewmembers not only work together to monitor the operating environment outside the locomotive, they also collaborate in planning activities, problem solving, and identifying and mitigating potential risk. A conductor is defined as the crewmember in charge of a train or yard crew. Freight conductors supervise pre-trip activities, over-the-road operation, and post-trip activities to ensure overall safe and efficient train movement.

The freight conductor's role has evolved from primarily a physical in nature job to one that emphasizes cognitive work. The research identifies five broad categories of cognitive job duties that a freight conductor normally faces, which raises issues for each railroad that might be considering one-person train operations and how the one-person operation can be as safe as a two-person operation.

One of those five categories of cognitive job duties is to manage the train consist, including the train makeup. This duty requires the freight conductor to understand train makeup rules and apply them both in the yard and on the mainline. Experienced conductors understand the implications of car placement, car consist, and car weight and shape when building trains. Conductors must understand how the train's consist will affect train handling, which is important to ensure locomotive engineer compliance when operating the train. (It is possible that this duty could also carry over to passenger train conductors, if there were different types of passenger cars in the same train that had the potential for compatibility issues,

e.g.,

incompatible doors.)

Second, a freight conductor also has the duty to coordinate with the engineer for safe and efficient en route operations, which includes checking speed, signal indications, and engineer alertness. This duty could also include filling an engineer's knowledge gap about a territory (

e.g.,

the conductor instructs the engineer where to place a train of a certain length so the train does not block a crossing). The conductor also serves to remind the engineer about upcoming signals and slow orders and provides “look ahead” information to alert the engineer about hills, curves, grade crossings, and other physical characteristics of the territory that have the potential to cause operational problems. If the locomotive engineer is not in compliance with the railroad's operating rules, it is the conductor's job to bring it to the locomotive engineer's attention, or take appropriate corrective action that may include actuating the emergency brake to bring the train to an emergency stop if the conductor feels the train, its crew, or others outside the train are in danger. A significant finding was that operating in mountain-grade territory adds complexity to the job and introduces additional cognitive demands on both the conductor and the locomotive engineer.

Third, a freight conductor's duties usually extend to taking the lead on interacting with non-crewmembers, such as dispatchers and roadway workers. These communications with non-crewmembers typically takes place by radio. There may be expected and unexpected radio communications, and there may be lulls in communication and times of heavy interaction that require conductors to multitask in order to simultaneously receive/copy information received by radio while calling out signals and speed restrictions.

Fourth, the freight conductor's duties require diagnosing and responding to train problems, as well as dealing with other exceptional situations.

Fifth, railroads typically assign the freight conductor the job of managing the train crew's paperwork. Examples of paperwork managed by a freight conductor include the conductor's log, writing down orders, copying bulletins for both crewmembers received by radio, and keeping an up-to-date rulebook. When a conductor is handling all of these duties, the safety benefit is that the engineer can concentrate on operating the train.

Another issue mentioned separately in this study's final report is that in order to gain the cognitive skill and knowledge to be an expert freight conductor, a person needs about 5 years of experience. This is because there are a significant number of overarching cognitive challenges that differentiate expert conductors from less experienced ones. A quick list of these overarching cognitive challenges include knowledge of the territory, the ability to maintain situational awareness of surroundings, the ability to project the effect of consist on train dynamics, the ability to problem-solve, the ability to plan ahead, the ability to multitask, the ability to exploit external memory aids, and the ability to foster situational awareness through active communication. The study concluded that less experienced conductors are less able to handle situations that require multiple demands on attention, and they are less able to effectively problem-solve, plan ahead, or identify and avoid potential hazards. Because they have had less “first-hand” experience on the job, they are typically less confident in their knowledge and ability. Having a two-person crew broadens the number of experiences from which the crew can draw from.

This research also addresses the role of PTC technology and whether it can substitute for a conductor, thereby paving the way for one-person operations. The cognitive task analysis addresses this issue by laying out the multiple ways in which conductors contribute to safe and efficient train operations and contrasts this with the anticipated features of PTC systems. The report concludes that PTC can provide warnings of upcoming signals, work zones and speed restrictions; however, PTC cannot account for all the physical and cognitive functions that a conductor currently provides. For instance, conductors can support locomotive engineers in monitoring events outside the cab window for potential obstacles and hazards undetected by automated systems (

e.g.,

people working on or around the track, trespassers, cars at grade crossings). FRA acknowledges that to the extent railroads comply with this rule using crewmembers in places other than the controlling cab, the crewmember is less likely to be able to provide this function. Other functions the conductor provides is filling knowledge gaps that locomotive engineers may have, supporting decision making, handling unanticipated events, and keeping the locomotive engineer alert, especially on long, monotonous trips where there is a risk of falling asleep. For this reason, the research recommends that each railroad seeking implementation of one-person operations in the future compile a detailed list of all of the physical and cognitive tasks both the engineer and conductor perform in the cab, determine which of these tasks PTC will cover, and understand how the locomotive engineer's responsibilities would change in a one-person operation. Of course, as the one-person crew would

presumably have more required tasks than an engineer in a two-person crew (even if PTC addresses some of those tasks), the railroad should consider how the strain of additional responsibilities may impact situational awareness. FRA requests comments on how railroads can and do safely and effectively perform these tasks using one-person crews.

Removal of the freight conductor from the most common arrangement of a two-person train crew team would have significant implications for the remaining one-person crewmember. One-person train crews would need to absorb the physical tasks necessary for operations, as well as the many cognitive tasks. Some of the freight conductor's current cognitive duties would be impossible with one person. For example, with a one-person crew, there will not be a second crewmember to fill in the knowledge or experience gaps of the sole crewmember. One of the problems is that inexperienced people “don't know what they don't know” and therefore cannot anticipate the risk and challenges, and cannot prepare for them. Pairing a conductor and locomotive engineer so that at least one of them is highly experienced can mitigate that problem.

Another potential issue of one-person crews is that it eliminates the opportunity to work as a conductor before promotion to locomotive engineer. This is a two-fold problem. First, engineers do not get the experience of separately learning the freight conductor position. Second, engineers who are never conductors are likely to begin their engineer careers with less railroad experience than those who first become conductors. Railroads that have used previously promoted conductors for their current one-person operations may find a shortage of such competent candidates to promote within the company if they eliminate the conductor position.

(2) “Rail Industry Job Analysis: Passenger Conductor,” Final Report, dated February 2013, DOT/FRA/ORD-13/07. The research and report was performed by the John A. Volpe National Transportation Systems Center and can be found online at

http://www.fra.dot.gov/eLib/details/L04321.

The purpose of this analysis was to identify key aspects of the passenger train conductor job, including the main responsibilities of the job, and the kinds of knowledge, skills, abilities, and other characteristics (KSAOs) required to successfully perform the job. The results of the analysis are useful to the railroad industry for three reasons. First, the results can be used to build training programs that address relevant and measurable KSAOs. Second, the results can be used to form the foundation for performance appraisal systems that are legally defensible and evaluate employees based on KSAOs that have been identified as related to the job. Third, the results can be used to help ensure that a hiring organization will appropriately screen new talent.

In relation to the crew size issue, this study is relevant because it explains the wide variety of KSAOs a passenger train conductor needs to possess in order to do the job well. Therefore, if a passenger railroad employs only a one-person train crew, there is a question of how one person can do all of these tasks and the tasks required of a locomotive engineer. Examples of passenger conductor KSAOs include knowledge of operating and safety rules, skill in working on and around moving equipment, judgment and decision-making ability, and a commitment to safety. Conductors use a number of different tools and types of equipment, and work with a variety of railroad personnel such as locomotive engineers, dispatchers, and foremen. The job is also physically and psychologically demanding for workers because of the prevalence of irregular work hours, out-of-doors work, and the need to lift and move heavy equipment. Passenger conductors also need to be able to carry out tasks involving passenger interaction; crew communication; crew supervision; form and record management; train inspection, troubleshooting, and repair; train makeup and handling; and emergency situations.

(3) “Fatigue Status in the U.S. Railroad Industry,” Final Report, dated February 2013, DOT/FRA/ORD-13/06. This report can be found online at

www.fra.dot.gov/Elib/Document/2929.

The research and report was performed by QinetiQ North America and an Engineering Psychologist within FRA's Office of Research and Development.

Train and Engine (T&E) workers, such as locomotive engineers and conductors, are safety-critical railroad employees that have the highest exposure to fatigue in the railroad industry. They are also among employees that have the longest work hours and work at night. Passenger T&E workers, as a group, are workers with the least fatigue exposure because of the predictability of their schedules and less nighttime work; however, some passenger or commuter workers are required to stay at an out-of-town location and do not return to their starting location at the end of the work period. Freight T&E work can be divided into two groups: (1) “road freight” work which involves moving trains over long distances between major terminals or interchange points and frequently requires overnight stays at an out-of-town location, and (2) “local freight” work which involves moving trains between a railroad yard and a nearby location so that the employee returns to the starting location at the end of the work period. Railroad workers are more likely to get less than seven hours of total sleep on a work day, which puts them at risk of fatigue.

Extrapolating from the findings in the study, it appears that a railroad considering a one-person train crew operation should consider whether the crewmember is likely to be fatigued. In a railroad's safety analysis, prior to implementing a one-person operation, it would be prudent for the railroad to consider what redundancy backstops have been implemented in case the crewmember falls asleep on the job. If FRA needed to review and approve an operation with less than two crewmembers, the agency would be looking to see if the railroad implemented strategies for reducing railroad worker fatigue, such as improving the predictability of schedules, considering the time of day it permits one-person train crews to operate, and educating workers about human fatigue and sleep disorders. This study could help provide a railroad with some ideas for reducing fatigue in its train crewmembers.

(4) “Technology Implications of a Cognitive Task Analysis for Locomotive Engineers—Human Factors in Railroad Operations,” Final Report, dated January 2009, DOT/FRA/ORD-09/03. The research and report was performed by the John A. Volpe National Transportation Systems Center and can be found online at

www.fra.dot.gov/Elib/Document/381.

This report documents the results of a cognitive task analysis (CTA) that examined the cognitive demands and activities of locomotive engineers in today's environment and the changes in cognitive demands and activities that are likely to arise with the introduction of new train control technologies. One of the objectives of this CTA was to understand these potential new performance demands. Another of the CTA's objectives was to evaluate the interaction between the locomotive engineer and the conductor and how they work jointly to operate the train in a safe and efficient manner. At the time of the CTA, the researchers assumed that railroads would continue to use a two-person crew configuration and so the analysis in this report does not explicitly consider any additional

sources of cognitive workload that may arise should there be a transition to single-person operations. The study notes that each crewmember has a duty to catch and correct the errors made by the other crewmember.

The research examined the following types of PTC systems: (1) Communications-based train management (CBTM), (2) advanced speed enforcement system (ASES), (3) incremental train control system (ITCS), (4) electronic train management system (ETMS), and (5) North American Joint Positive Train Control (NAJPTC). This 2009 study acknowledges that the PTC systems are described and analyzed as they were implemented at the time of the site visits and, in some cases, the PTC systems may have undergone substantial redesign since then.

The results pointed to major cognitive challenges involved in operating a train, including the need for sustained monitoring and attention; maintaining an accurate situation model of the immediate environment (including the location, activities and intentions of other agents in the vicinity such as other trains and roadway workers); anticipating and taking action in preparation for upcoming situations; and planning and decision-making, particularly in response to unanticipated conditions (

e.g.,

person or object obstructing the track). Introduction of new train control technology reduces some cognitive demands while creating new ones. For example, as four out of the five PTC systems tested used conservative braking profiles to slow the train to the desired target speed under restrictive assumptions (

e.g.,

heavy train or slippery track), train crews discovered that they would need to initiate braking at an earlier point than they were normally accustomed to if they wanted to prevent the PTC system from braking the train for them. This earlier braking point conflicts with the experienced crews' effective strategies for operating as efficiently as possible. A penalty brake application is highly undesirable because it significantly delays train operations and may trigger report or documentation requirements to explain why the penalty brake occurred. The report also discusses the implication of the results for design of in-cab displays and development of training, particularly for PTC systems. The research suggests there is a need for development of in-cab displays that make it easier to anticipate and stay within the braking curve without having to look closely at the in-cab display so that more attention can be directed to looking outside the window.

The PTC systems also created new sources of workload and distraction. Sources of workload and distractions include the need to acknowledge frequent (and often non-informative) audio alerts generated by the PTC system and the need for extensive input to the PTC system during initialization and when error messages occur while operating the train. For example, the NAJPTC system is described as having a train location determination system (LDS) that is able to locate train position within 10 feet but it would trigger a failure alarm when the LDS system experienced difficulty identifying the train location. The failure alarm sounded repeatedly, requiring the train crew's attention. Although this situation described was an early test of the system, and no consequences of failing to respond to the alert occurred, when the test period ends a failure to respond to an alert quickly might result in a penalty brake. The experiences of European railroads suggest that the concern expressed by the locomotive engineers regarding too many non-informative alerts has a potential for negative safety consequences. Operators may respond to poorly designed audio alerts automatically without fully processing their meaning, thus defeating their purpose. This is consistent with an extensive body of human factors literature that indicates that individuals are likely to ignore alarms when a high false alarm rate exists. (Please note that FRA's PTC regulation prohibits requiring a locomotive engineer to “perform functions related to the PTC system while the train is moving that have the potential to distract the locomotive engineer from performance of other safety-critical duties,” which would include distracting, non-useful alerts.

See

49 CFR 236.1006(d)(1), formerly § 236.1029(f)).

The new cognitive demands created by new technologies such as PTC can lead to changes in how locomotive engineers operate the train. Locomotive engineers certainly combine the current information they can obtain from direct perception (

e.g.,

displays inside the cab as well as the scene outside the cab), in addition to knowledge and skills gained through training and experience to develop train handling strategies. Sources of new cognitive demands include constraints imposed by the PTC braking profile that require locomotive engineers to modify train handling strategies, increases in information and alerts provided by the in-cab displays that require locomotive engineers to focus more attention on in-cab displays versus out the window, and requirements for extensive interaction with the PTC systems (

e.g.,

to initialize it and to acknowledge messages and alerts) that impose new sources of workload. The research concluded that although PTC technology is likely to have a positive impact on overall risk of accidents, these new sources of cognitive demand can contribute to errors and accidents.

Railroads and PTC system designers need to be made aware that measures can be taken in the design of PTC displays and in development of user training to improve train crew performance and reduce the potential for human error. The final section of this report discusses a number of suggestions for ways to improve in-cab displays to reduce cognitive demands on train crews and facilitate train crew performance as well as suggestions for improved training. For example, one promising area for research and development is improved in-cab displays that minimize the need to visually attend to the in-cab display to extract important information. The research found that a substantial learning curve exists to reach the point where the in-cab display does not serve as a source of distraction, diverting attention away from events out the window. Locomotive engineers must have sufficient experience in running a PTC-equipped train as part of training so that they get beyond the point where close monitoring of the in-cab display is required to avoid a penalty brake application.

Another PTC issue related to crew size is that PTC systems generally require manually entered inputs at the start of a trip and after a shutdown of the system during train operations. The train crew must enter information that the system will use as parameters for safe operation. These data entry tasks provide another source of workload and distraction, yet they are highly important because manual entry errors can have safety implications. With a one-person crew, the task burden would fall on the sole crewmember. Although a railroad might consider that if there is only one-person in the locomotive cab, the person should not operate without the PTC system operational, reinitializing the PTC system after it has initiated a penalty brake application can be a complex and time-consuming procedure. On one railroad described in the research, the procedure is so complex, difficult to follow, and time-consuming that, during the PTC system's trial period, the locomotive engineers were allowed to forego reinitializing the PTC system. However, the study noted that once the system

becomes fully operational, running a train without PTC activated may no longer be an option.

This study is important to the crew size issue because it challenges the possibility that a train with PTC is inherently safe with only a one-person crew and that no safety analysis or FRA oversight of the operation is warranted. The study concluded that although PTC technologies have the potential to improve safety and efficiency of railroad operations, they also have the potential to create new failure modes and impose new cognitive demands on locomotive engineers who need to monitor PTC displays and provide inputs to the system. For example, without PTC technology, locomotive engineers are highly engaged with the train operation, noticing visual cues (

i.e.,

landmarks and mileposts), monitoring radio communications of other trains, and relaying information by radio to other trains about potential hazards. Some locomotive engineers even indicated that they get a variety of sensory-based cues that help them perceive their location, such as vibrations associated with a portion of track or a smell that reminds them they are near a farm. The research suggests that because the PTC technology may require locomotive engineers to focus more of their attention on in-cab displays, it will reduce their ability to monitor activity outside the cab and raises a question about whether the engineers will lose any situational awareness in relation to the coherent mental picture (

i.e.,

the situation model) of where the engineer perceives the train to be based on prior experience. Typically, a locomotive engineer will use that situation model to help the engineer anticipate future events. Furthermore, the research concluded that train crews must avoid too much reliance on the new train control technologies because, if the system ever fails, the engineer must still be able to operate the train safely.

(5) “Using Cognitive Task Analysis to Inform Issues in Human Systems Integration in Railroad Operations-Human Factors in Railroad Operations,” Final Report, dated May 2013, DOT/FRA/ORD-13/31. The research and report was performed by the John A. Volpe National Transportation Systems Center and can be found online at

http://www.fra.dot.gov/eLib/details/L04589.

Human Systems Integration (HSI) is defined as a systematic, organization-wide approach to implementing new technologies and modernizing existing systems that emphasizes the importance of the end-user in the system acquisition process. FRA sponsored this research because it would like the railroad industry to consider HSI when implementing new technologies such as PTC, energy management systems (EMS), and electronically controlled pneumatic (ECP) brakes in the locomotive cab. The expectation is that an HSI approach to railroad technology acquisition and implementation can increase user acceptance and usability of the technology, as well as increase the likelihood that it is deployed successfully. This report provides guidance to the industry with respect to the need for HSI in the technology acquisition process, and more specifically, how to use Cognitive Task Analysis (CTA) methods and results as part of the HSI process.

The nature of the work associated with many railway worker positions (

e.g.,

locomotive engineers, conductors, and roadway workers) is rapidly shifting from being primarily physical to placing greater emphasis on cognitive demands (

e.g.,

monitoring, supervising automated systems, planning, communicating and coordinating, and handling unanticipated situations). CTA methods provide a means to explicitly identify the knowledge and mental processing demands of work so as to be able to anticipate contributors to performance problems (

e.g.,

lack of information, high attention demands, inaccurate understanding) and specify ways to improve individual and team performance (be it through new forms of training, user interfaces, or decision-aids). CTAs can inform all aspects of HSI starting from early system requirements exploration and definition through late stage validation and field testing. The information in the report can serve as a lead-in to the kinds of insights that can be drawn from performing a CTA when introducing new technologies into railroad operations, as well as a starting point for the industry as far as identifying the likely emerging issues that need to be explored as a result of the introduction of new technology. For example, CTA methods can examine how the introduction of PTC might impact the monitoring demands placed on locomotive engineers, or alter the patterns of communication between locomotive engineers and other railroad workers. CTA methods can inform the design of systems that are more likely to be successful when deployed by ensuring that they address the specific performance challenges users face and are sensitive to the larger system context. A CTA can be used to better understand the various roles and responsibilities associated with each crew position to be able to assess which of those roles and responsibilities are eliminated (or taken on) by the new technology and which remain and must be accounted for in some other way if the crew position is eliminated. FRA has significantly aided this HSI analysis by previously sponsoring CTA reports that focused on railroad dispatchers, roadway worker activities, locomotive engineers, and freight train conductors (the two latter reports were previously described in this preamble section).

The report cites a prior research finding that the introduction of new technology does not necessarily guarantee improved human-machine system performance. Woods, D. & Dekker, S., “Anticipating the effects of technological change: A new era of dynamics for human factors,” Theoretical Issues in Ergonomics Science, 1(3), 272-282 (2000); National Research Council (NRC) Committee on Human-System Design Support for Changing Technology, “Human-System Integration in the System Development Process,” National Academies Press (2007),

http://www.nap.edu/catalog.php?record_id=11893;

and Wreathall, J., Woods, D.D., Bing, A.J. & Christoffersen, K., “Relative risk of workload transitions in positive train control,” Washington, DC: U.S. Department of Transportation, Federal Railroad Administration. DOT/FRA/ORD-07/12 (2007),

http://ntl.bts.gov/lib/42000/42400/42472/ord0712.pdf.

Poor use of technology can create additional workload for system users, can result in systems that are difficult to learn or use, or, in the extreme, can result in systems that are more likely to lead to catastrophic errors. The introduction of new technology results in the following types of common changes in operating practice: (1) Changes in practitioner roles, including emergence of new tasks; (2) changes in what is routine and what is exceptional; (3) changes to the kinds of human errors that can occur; and (4) people in their various roles adapting by actively altering tools and strategies to achieve goals and avoid failure. HSI is a way to employ a comprehensive analysis, design, and evaluation process that mitigates the risk of designing systems that create potential mismatches between the technology and the human operator limitations or capabilities. For example, in reviewing the freight train conductor CTA and how it could inform the HSI process regarding issues of one versus two-person train crew operation, the study concluded that “[i]t is not clear how the introduction of PTC will affect cognitive and collaborative

processes, but findings suggest that it will not account for all the cognitive and physical support functions the conductor currently provides.”

The study found that there are other CTA methods that can be used to provide more fine-grained input to HSI analysis and design activities. For example, there are CTA methods that provide a more detailed, second-by-second description of the mental processes (

e.g.,

perceptual processes, attention processes, memory store and retrieval processes) involved in performing complex cognitive tasks such as operating a train. The study provides descriptions and citations to these recent attempts to examine the microlevel (second-by-second) information processing involved in operating the train over a route. These more microcognitive-level analyses can be particularly helpful for analyzing attention and workload demands at an in-depth level.

In the emerging issues section of the report, the study explained that if a railroad chooses to transition to one-person operations based on technology such as PTC, a proper HSI analysis would require that the railroad answer certain fundamental questions about the operation for the system designers. For instance, will the engineer still be responsible for manually operating the train? If not, when will the engineer manually control the train? When will the software (automation) system operate the train with the engineer acting as supervisor? And, when will the roles be blended? Answers to these questions may introduce additional concerns. For example, situational awareness and operator vigilance may become more of a concern when the engineer's role becomes more supervisory. If crew size is reduced to one person, how will the reduction in crew size impact safety when the one-person crew is used to relying on cooperative strategies with the second person that fosters shared situational awareness and creates safety nets?

(6) “Teamwork in U.S. Railroad Operations,” A Conference, April 23-24, 2009, Irvine, California, Transportation Research Board, Number E-C159, dated December 2011. The many authors of the research and reports are listed in the publication which can be found online at

http://onlinepubs.trb.org/onlinepubs/circulars/ec159.pdf.

This conference report discusses the key aspects of successful teams, such as train crews. The Transportation Research Board is a division of the National Research Council, and an independent adviser to the Federal government and others on scientific and technical questions of national importance. This particular conference drew upon the expertise of researchers and operating personnel concerned with human performance and human factors research issues related to railroad operations. The following is a summary of some of the relevant discussions in the conference report. The report contains citations to the research each presenter relied on in forming their analyses and conclusions.

One central theme is that teams do not become expert without guidance. They must be trained according to the established scientific principles. But training alone is not enough. To facilitate its success, organizations must promote and reinforce teamwork behaviors. Long-term organizational commitment is crucial to demonstrating that teamwork training is not just a fad, but is a central component of company policies and procedures. In other words, there needs to be a “culture of teamwork” embedded within the organization.

Team performance can be improved when members effectively communicate. One effective example is when crews use periods of low workload to plan ahead, so that if a difficult situation arose, the explicit discussions become the basis for actions. Of course, a question implied from this report is that if the train crew consists of only one person, can the lone crewmember plan ahead during periods of low workload to the same extent as a crew of two or more persons who understand how to effectively communicate? Unfortunately, the conference report does not answer this implied question.

There are five critical components of teamwork: Mutual performance monitoring, backup behavior, adaptability, team leadership, and team orientation. Although not addressed by the conference report, arguably three of these strengths of teamwork are lost when the team consists of only one person. Team orientation refers to a person's tendency to prefer working with others, which could certainly be problematic if a person with a team orientation is ordered to operate a train as a one-person team. Mutual performance monitoring refers to the ability to keep track of fellow team members' work while carrying out their own, to ensure that everything is running as expected, and to ensure that they are following procedures correctly. Mutual performance monitoring is necessary in teams in order to prevent teams from making errors and enable teams to engage in backup behaviors. Backup behavior occurs when a team member recognizes that another team member is in need of aid and offers assistance. Backup behavior requires team members to know enough about other team members' responsibilities to anticipate their needs. Research has identified three types of backup behavior: (1) Providing feedback to improve performance, (2) assisting a teammate in performing a task, and (3) completing a task for a team member who is overloaded. The benefits of mutual performance monitoring and backup behavior are simply lost when the team consists of a single employee.

One comment FRA heard during the RSAC Working Group meetings was that multiple person train crews could be less safe than a one-person crew because sometimes crewmembers distract each other from the train operation activities. This issue was addressed in the conference report with regard to a discussion of how expert teams perform versus non-expert teams. An example was given of a train accident in which a student engineer was allowed to operate a train independently, receiving no guidance through supervisor role modeling or feedback prior to a collision. The incident was an exemplary prototype of a non-expert team because not only were the crewmembers not trained adequately with effective feedback prior to the day of the accident, but also communication and coordination completely broke down between all team members directly before the incident. In contrast, expert teams have a clear and common purpose, as well as an understanding of each individual member's roles. It is that understanding that allows expert team members to anticipate each other's actions and back each other up when needed, as well as coordinate without explicit and lengthy communication. Furthermore, unlike non-expert teams, expert teams engage in a regular cycle of prebrief, performance, and debrief. This performance cycle engages the expert teams to identify high and low priorities, revise goals and plans, identify lessons learned, and evaluate whether the team is or is not effective both in performing the task and identifying the needs of team members. The research in the conference report concludes that the main advantage of developing expert teams is that they have higher levels of performance. For example, expert teams make better decisions and fewer errors, which in turn enable expert teams to have a higher probability of mission success.

In yet another of the presentations in the conference report, an issue raised was whether internal and external

communications of train crewmembers could be captured to consider the impact of new technology, such as PTC, on crew interactions and performance. The report states that making the most of new technologies to improve efficiency while maintaining safety and augmenting effectiveness will always present challenges, but that prudent application of team science in general and of communications analysis in particular can both facilitate their achievement and enhance their utility. The report explains that those technologies place new demands on train crews in terms of tasks to be performed, skills required, and the size and mix of both onboard and distributed teams. FRA notes that, based on RSAC Working Group discussions, some railroads appear ready to reduce train crew size from two persons to one, upon implementation of PTC, under what FRA and the presenters of this report suggest would be a wrong presumption that with PTC there would be less tasks for the crew to do or the tasks would be easier to accomplish with a single person. The report counters that presumption and suggests that the impact is unknown until PTC is implemented and the impact it would have on a two-person crew is studied.

C. The Acknowledged Limitations of FRA Accident/Incident Reporting Data

FRA's accident/incident data is derived from the agency's requirements for railroads to record and self-report specific information to FRA. The purpose of FRA's accident/incident recordkeeping and reporting regulation, contained in 49 CFR part 225, is “to provide the Federal Railroad Administration with accurate information concerning the hazards and risks that exist on the Nation's railroads. FRA needs this information to effectively carry out its statutory responsibilities under 49 U.S.C. chapters 201-213. FRA also uses this information for determining comparative trends of railroad safety and to develop hazard elimination and risk reduction programs that focus on preventing railroad injuries and accidents.” 49 CFR 225.1. Over the life of the part 225 regulation, FRA has amended these requirements in an effort to require railroads to improve the accuracy of their reporting.

See

75 FR 68862, 68863-64 (providing an overview of part 225 and recent amendments). FRA does not investigate every reportable accident/incident, but frequently conducts audits and investigations to ensure that railroads are accurately reporting. In 2013, FRA conducted its own investigation of 89 train accidents/incidents that FRA determined might have greater significance to the industry or the general public. FRA did not have the resources to investigate all of the 1,781 train accidents/incidents railroads reported to FRA in 2013. FRA is not aware that any of the accidents/incidents it investigated involved a one-person crew operation.

Part 225's central provision requires that each railroad subject to part 225 submit to FRA monthly reports of all accidents and incidents that meet FRA's reporting criteria. 49 CFR 225.11. Railroad accidents/incidents are divided into three groups, each of which corresponds to the type of reporting form that a railroad must file with FRA: (1) Highway-rail grade crossing accidents/incidents (FRA Form F 6180.57); (2) rail equipment accidents/incidents (FRA Form F 6180.54); and (3) deaths, injuries and occupational illnesses (FRA Form F 6180.55a).

See

49 CFR 225.19. For the reporting of deaths, injuries, and occupational illnesses that result from an event or exposure arising from the operation of a railroad, the FRA forms do not request that the railroad record the number of crewmembers as that distinction is unlikely to be pertinent to accident analysis for those types of accidents/incidents; instead, FRA only requires that the railroad report which crewmembers were injured, killed, or suffered an illness. Thus, it is impossible to search FRA's accident/incident database for those forms to find whether a death, injury, or occupational illness did arise from the operation of a train with a one-person crew. Meanwhile, for the first and second group, highway-rail grade crossing accidents/incidents and rail equipment accidents/incidents, the FRA forms record the number of crewmembers. The highway-rail grade crossing accidents/incidents form records the number of people on the train at the time of the accident (both passengers and train crew). The rail equipment accidents/incidents form records the number of crewmembers in boxes 40-43, with four different work positions listed: Engineer/Operator, Fireman, Conductor, and Brakeman. Obviously, FRA does not see as many Fireman and Brakeman listed as it once did, but they are still occasionally listed. The railroad must record the number of each type of crewmember that was working on the train at the time of the accident/incident. Thus, FRA is able to search the records to determine how many train crewmembers were assigned to a train that was involved in a reportable rail equipment accident/incident or a grade crossing accident.

FRA is considering including in the final rule a requirement to report train crew size data in the deaths, injuries, and occupational illnesses accident report form. Such a regulatory change would allow FRA to have crew staffing information and to better assess the performance of train crews with less than two members. The benefits of this proposed change would be evaluated while FRA conducts a future comprehensive reform of its accident/incident reporting forms to modernize and meet data needs. As it relates to crew staffing and its characteristics, the impetus for this effort originated during the RSAC Working Group meetings regarding train crew size. This effort made it clear that there is a need to improve both the quality and the scope related to the collection of information of train crew staffing safety. As presented above, existing data forms do collect information about the number of crewmembers involved in a train accident. However, current reporting requirements do not provide all the information required to assess the safety performance of crews with less than two members. Likewise, FRA data needs outside of this rulemaking are numerous and need to be contemplated. For these reasons, FRA is engaged in an effort to review and determine what data collection practices need to be changed. However, FRA also concluded that this effort has to be thoughtful and broad to ensure it collects high quality data. FRA is considering how to prioritize items and decide what data to collect on items such as ECP brakes, PTC, or crude oil or ethanol transportation by rail. All these matters are of high priority and would have to be considered in a comprehensive manner to minimize information collection burden on the regulated community. This NPRM is useful to request public input as it pertains to crew staffing data and determine what type of information collection needs to be refined or what clarification in the part 225 guidance needs to be amended to ensure forms are completed correctly. This input would be used to inform a future rulemaking that would propose changes to part 225, FRA Form F 6180.54, and its related guidance.

For the benefit of the RSAC Working Group, FRA reviewed nearly 12 years of railroad safety data between January 2002 and October 2013 by searching the F 6180.54 rail equipment accidents/incidents forms. FRA manually reviewed 1,443 reports and applied several filters to eliminate redundant reports, other than human-factor caused

accidents/incidents, accidents/incidents that occurred within railroad yards, and accidents/incidents involving railroad maintenance equipment. After applying these filters, FRA was left with accidents/incidents that railroads informed FRA were caused by human error and involved a one-person crew operating on main track. The result of this review was that FRA identified 28 human-factor caused accidents/incidents involving one-person crews operating conventionally and four accidents/incidents involving remotely controlled operations on main track. Since FRA does not capture data that would provide information regarding the total operating mileage for one-person crew operations in the United States (or even two-person operations), it is impossible for FRA to normalize the data and be able to compare the accident/incident rate of one-person operations to that of two-person train crew operations to see if one-person operations appear safer or less safe. Additionally, one-person operations over this period are not constant and use of one-person train crews for operations on main track appear to be increasing over the past several years, so there are additional factors that could make historical rates less of an indicator of current or future rates.

The accident/incident reports involving one-person train crews also do not clearly help determine that the accident/incident would have been prevented by having multiple crewmembers. FRA requires railroads to determine the primary cause of a rail equipment accident/incident and enter a primary cause code on the form. If possible, railroads are also encouraged to enter a contributing cause code on the form as well. FRA does not have a cause code that a railroad could use to indicate that a one-person train crew caused the accident. In other words, there is no cause code that directly suggests that the reporting railroad believes the accident/incident could have been prevented by having a second crewmember. Even if FRA were to add such a code, a railroad would have a disincentive to use it as doing so might suggest that the railroad employ more crewmembers, increasing wage costs. Of course, if a railroad thought that only having one person was a factor, FRA has a cause code, M599, that may be used when no other cause codes apply. If M599 is used, the railroad must describe the events in a narrative. Furthermore, FRA relies on each railroad to self-report a description of the accident/incident, as well as the primary and contributing causes. Without an accurate description and identification of the causes, FRA personnel reviewing the report might not believe there is the potential that a second person could have helped prevent the accident/incident.

After RSAC failed to reach consensus, FRA conducted additional accident/incident data searches in an effort to determine whether there were any trends that could be identified. FRA looked at whether any data might have suggested a safety problem with MMA, which operated the train in the tragic Lac-Mégantic accident described earlier, or with any problems with shortline railroads that were similar in size to MMA. Rather than compare MMA to the entire railroad industry which could provide a distorted result (as just a few accidents on a shortline might make it look like it has a high accident rate compared to a major railroad that operates many more miles over the course of a year), FRA compared MMA to its shortline peers. In 2012, the last full year before the accident, MMA had about 160,000 total miles. FRA reviewed its accident/incident database from 2003 through April 2014 and compared MMA to the 52 other railroads that had total miles in 2012 of between 100,000 and 200,000. FRA also looked at the data to see if it could determine the number of accidents for each of these shortlines, with and without one-person crews. For the one-person crews, FRA was able to isolate train accidents where hazardous materials were in the train, and eliminate remote control operations and any operation that occurred on yard track.

The data concerning MMA and its shortline peers revealed that nearly half of the 52 shortlines (25, or 48 percent) had at least one accident where hazardous materials were in the train, but that MMA had the worst record in this category. MMA had 18 accidents, which was twice as many as its closest shortline peer. MMA's 18 accidents accounted for 23 percent of the 78 total number of accidents in its shortline peer group where hazardous materials were in the train. Although only 4 of these 78 accidents/incidents occurred with a one-person crew (about 5 percent), 2 of the 4 occurred on MMA. Looking at all one-person crew train accidents in which a MMA shortline railroad peer reported the cause to be a human factor failure, MMA reported no such accidents and 9 of MMA's shortline peers reported a total of 13. Consequently, while it can be determined that the two MMA one-person crew accidents involving hazardous materials in the train were not reported by MMA to be caused by a human factor failure, the data suggests that MMA stood out as having significantly more accidents involving trains carrying hazardous materials than its peers.

When looking at all train accidents in which a MMA shortline railroad peer reported the cause to be a human factor failure, MMA reported four such accidents, 4 of MMA's shortline peers also reported 4 such accidents, 13 of MMA's shortline peers reported more than 4 such accidents, and 39 of MMA's shortline peers, including MMA, reported a total of 153 human factor failure caused accidents. Including MMA, over 70 percent of MMA's shortline peers had at least one train accident caused by human factor failure, and 25 percent had more human factor failure train accidents than MMA. Thus, MMA did not stand out among its peers as having a much higher number of accidents attributed to human factor failure. FRA believes that even in cases where problematic one-person train operations cannot be identified by their number of past human factor accidents, FRA would be able to identify such operations with other information including inspection reports, and the railroad's description of operations and contingency plans to evaluate the safety culture and overall emergency preparedness to handle one-person operations.

If FRA were only to focus on the one-person crew safety data prior to the Lac-Mégantic accident, it would have been difficult to make the case that MMA did not have a good enough safety record to operate one-person train crews as MMA did not have any accidents/incidents that it attributed to human factor failure of the one-person train crew. It also only had 2 one-person crew accidents involving hazardous materials in the train over the more than 10-year period analyzed. However, if this NPRM is finalized, FRA could use the data suggesting MMA had significantly more accidents involving trains carrying hazardous materials than its peers to have MMA address safety issues to reduce the overall high number of accidents before providing FRA approval of the continuance of a one-person train operation or approval for a new one-person operation.

See

49 CFR 218.133 and 218.135.

Furthermore, this is an example of when the limitations of FRA's safety data would not help make a direct case that one-person operations are less safe than multiperson train crews but may still provide some possible basis for this proposed rule. That is, FRA's safety data suggests that a particular railroad that has a higher rate of train accidents

where hazardous materials are in the train could find itself more likely to continue that trend regardless of the size of the crew, assuming the railroad takes no action to further prevent such accidents from occurring. And if such accidents were to eventually occur, FRA has found that multiperson train crews are better equipped to protect each other, other railroad workers, railroad equipment, the environment, and the general public, because they have more options available to them for taking mitigation measures than a single crewmember. Thus, a derailment might occur, regardless of the number of train crewmembers, but it might be the actions of the train crew post-accident that determine the severity of the damages or injuries that result. This may be especially so when hazardous materials are present in the train or are in other trains operating on the same or adjacent track.

While data and information about one-person operations around the world are limited, evidence found by FRA and explained in the Regulatory Impact Analysis (RIA) that accompanies this rulemaking indicates that the safety records of these foreign operations are acceptable. FRA also found that most of these foreign operations would meet the requirements in one of the exceptions of the proposed rule (due to their size), and that most foreign governments have a role in the implementation of one-person crews (where they exist). Another factor to consider is that railroad workers in other countries have a more predictable work schedule, fewer working hours per week, and more opportunities to rest.

See

RIA Table 4. Nonetheless, FRA requests public comment on the lessons learned from these nations to implement one-person crews under a balanced regulatory oversight. Additionally, FRA requests public input about the safety performance of passenger and freight rail operations with less than two people in other countries. This is important because FRA could not find specific data on the safety records of international one-person crew passenger operations that do and do not meet the proposed exceptions.

Finally, railroads have achieved an improving safety record during a period in which the industry largely employed two-person train crews. FRA has no empirical evidence to suggest a causal relationship between these variables rather than a correlative one. In fact, it is possible that one-person crews have contributed to the improving safety record. Comparing calendar year 2004 to 2013, total accidents/incidents are down over 21.5 percent and human factor-caused train accidents/incidents are down over 50 percent. Over that same period, the number of reportable train accidents/incidents has decreased from 3,385 in 2004 to 1,781 in 2013, a decrease of over 47 percent. The normalized frequency index of 2.380 per one million train miles for 2013 represents the safest year in that 10-year period, and is a decrease of nearly 46 percent from 2004. Meanwhile, it is impossible to keep data on how many accidents/incidents were prevented by having a properly trained two-person crew, where each crewmember understood each other's duties and together could perform as an expert team. Thus, although the limitations of the data collected make it difficult to make a straightforward finding that one-person operations are more or less safe than two-person operations, FRA's approval process in this NPRM is expected to provide some insight into exposing dangerous operations and lead to safety improvements for those railroads that want to reduce the number of train crewmembers to less than two.

D. FRA's Regulations Were Designed for at Least Two Crewmembers

During the Working Group's first meeting, FRA presented the agency's position that many of the Federal rail safety regulations were written with the expectation that each train would have multiple crewmembers. That does not mean that FRA expects that at least two crewmembers will be in the cab of the controlling locomotive at all times, which may surprise some people who are not familiar with a wide-variety of railroad operations. A typical freight locomotive is founded with the expectation that multiple crewmembers could be working in the cab of the controlling locomotive. However, there are many operating circumstances in which a second crewmember could more effectively safeguard the operation by being somewhere other than the locomotive cab of the controlling locomotive and it would be difficult for a one-person train crew to perform the same operation. Because a railroad's operating rules and practices for a one-person operation will be a bit different than for multiple person train crews, some safeguards will be lost and new methods of operation will be developed to try and plug any regulatory holes. Without a crew size regulation, railroads would be free to jettison certain requirements that apply to multiple person crews without specifically being required to fully consider the potential safety repercussions. The following background explains some of the Federal rail safety requirements that will not work as intended when one-person train crews are deployed.

1. Difficulty Providing Point Protection for Shoving or Pushing Movements

For shoving or pushing movements, a second crewmember routinely provides point protection where the controlling locomotive is the furthest car in the train from the leading end.

See

49 CFR 218.99. In that case, a second crewmember riding the leading end or being on the ground in radio communication with the train's locomotive engineer may be the safest practice. A one-person train crew, operating any train of a significant length, may have difficulty determining that the track is clear for the shoving or pushing movement without the assistance of another person. Shoving blind,

i.e.,

not protecting the movement, would violate the Federal rule.

Passenger and commuter locomotives do not always have room for a second crewmember in the locomotive control compartment, but a second person may still be necessary to provide assistance for shoving or pushing movements. Pushing or shoving movements are routine operations and thus FRA's expectation is that few trains could perform these movements safely with only a one-person crew. We note, however, that the point protection rule permits use of cameras for performing these movements.

See

49 CFR 218.99(b)(3)(i).

2. Complications Returning Switches to the Normal Position and Loss of Job Briefings

In a typical multiple crewmember operation, the locomotive engineer would rarely be expected to leave the cab of the controlling locomotive to perform operational work. However, in a one-person operation, unless all switches can be operated from the locomotive or by a non-crewmember in accordance with a railroad's operating procedures, the locomotive engineer would encounter logistical difficulties in throwing some switches and then returning those switches and locking them in the normal position after use.

See

49 CFR 218.103 through 218.107. If the one-person crew were to throw the switches and return them to the normal position, the person would need to walk back and forth the length of the train each time a switch was returned to the normal position.

The Federal regulations concerning throwing switches anticipate that the crewmembers will conduct job briefings “before work is begun, each time a work plan is changed, and at completion of

the work.”

See,

49 CFR 218.103(b)(1). The regulation does not anticipate that a train crew consisting of one-person would be exempt from the job briefing requirements, although it seems absurd to think that any one-person train crews would need to hold job briefings with themselves. However, one of the most important benefits of a job briefing, with each crewmember's input, is potentially lost when there is a one-person operation. That is, a lone crewmember cannot benefit from another crewmember's experience about the best way to safely perform the operation. Under routine operations, one-person crewmembers will decide for themselves how best to proceed. The one-person crewmember will also assess the factual circumstances of a situation by themselves, without the benefit of any additional crewmembers' observations. Although a railroad could implement procedures to address certain types of operations that can aid a one-person crew, such a briefing may not be able to duplicate all of the information that a fellow crewmember could.

3. Concerns Protecting Train Passengers in an Emergency

During the first Working Group meeting, FRA made a presentation regarding FRA's passenger train emergency preparedness rule (49 CFR part 239) and explained how multiple train crewmembers are typically necessary in order to fulfill the purpose of the rule. The purpose of the passenger train emergency preparedness rule “is to reduce the magnitude and severity of casualties in railroad operations by ensuring that railroads involved in passenger train operations can effectively and efficiently manage passenger train emergencies.” 49 CFR 239.1(a). There are numerous ways that crewmembers, other than the locomotive engineer, can assist the passengers in an emergency. Emergencies can require evacuations in various types of circumstances where a trained person would be helpful to guide passengers away from danger. For example, passengers that self-evacuate might not realize that they could step on an electrified rail or be struck by a train approaching on an adjacent track. Evacuations in remote areas, in tunnels, or on bridges also pose significant dangers to passengers and are places where crewmembers are required to be trained on safe methods to assist passengers. A one-person crew would have significant difficulty coordinating any type of evacuation, especially in difficult terrain, if the crewmember cannot walk from car to car, or if there are large numbers of passengers. Furthermore, although signs for train passengers can be useful, signs have limited value for reliably instructing passengers on when it is safe or unsafe to evacuate under all conditions.

4. Deterrence of Electronic Device Distraction and Observing Alcohol or Drug Impairment; Reduced Possibility of Co-Worker Referrals

Another issue that could be a concern with a one-person train crew is whether there is adequate supervision to determine that the person is not reporting for duty under the influence of or impaired by alcohol or drugs. With multiple train crewmembers, a second crewmember might suspect that a person has used, or is using or possessing alcohol or drugs on railroad property. Working with a potentially impaired co-worker is a safety hazard that puts other crewmembers in direct conflict with one another. For that reason, FRA has developed minimum standards for co-worker report policies that allow the employee suspected of abuse to get treatment and rehabilitation, with the potential to return to railroad safety-sensitive work under certain conditions.

See

49 CFR 219.405 and 219.407 (permitting a railroad to implement an alternate co-worker policy with the written concurrence of the recognized representatives of a particular class or craft of covered employees). The co-worker referral policy makes it more palatable for an employee to turn in a potentially impaired co-worker, knowing that the co-worker will have an opportunity to get professional help without the co-worker necessarily losing his or her job, and not having to work side-by-side with that impaired co-worker.

Although a one-person crew may be subject to pre-employment testing, random testing, and testing for cause, each of these types of tests do not apply to shortline railroads which have a total of 15 or fewer employees who are covered under the hours of service laws and do not operate on the tracks of any other U.S. railroad. Additionally, even if a one-person crew is potentially subject to each of those tests, the person will not be tested before, during, or after every tour of duty. Thus, a one-person crew has more opportunity, especially on the smallest shortline operations, to conceal a drug or alcohol violation, than the person would if there were two or more crewmembers.

Similarly, without a second crewmember to monitor the sole crewmember's attentiveness, there is a risk that more locomotive engineers will be tempted to use cell phones and other prohibited electronic devices when nobody is around to observe them. When FRA issued a final rule restricting railroad operating employees from using cellular telephones and other electronic devices, FRA noted that distracted driving impacts all transportation modes because these devices have become ubiquitous in American society.

See

75 FR 59580, 59582, Sep. 27, 2010, promulgated at 49 CFR part 220, subpart C. In the justification for the rulemaking, FRA stated that it discovered numerous examples of the dangers posed by distracting electronic devices and described five rail accidents indicating the necessity for the restrictions. FRA's electronic device distraction rulemaking also stated that “it is difficult to identify distraction and its role in a crash” if it goes unreported by the operator of the vehicle. 75 FR at 59582 (describing how data on the number of motorcoach crashes may potentially understate the true size of the problem because “self-reporting of negative behavior, such as distracted driving, is likely lower than actual occurrence of that behavior). Thus, a second crewmember could act as both a deterrent to any crewmembers using electronic devices in a prohibited manner and as a witness reporting such inappropriate electronic device usage during an accident/incident investigation.

5. Complicating Radio Communication Procedures

Some radio and wireless communication requirements were written with the expectation that there would be at least two crewmembers on a train. For example, FRA requires that an employee copying a mandatory directive received by radio transmission shall not be an employee operating the controls of moving equipment.

See

49 CFR 220.61. Copying a mandatory directive would clearly be distracting to a person who was attempting to operate a train simultaneously, which explains why it is strictly prohibited. Certainly, a one-person train crew could stop a train to receive a mandatory directive by radio, but there is a question whether railroads have thought through all the safety implications of stopping the train. The train may be going at a high enough speed that it would take over a mile to stop the train, or the train might be in a territory where a steep grade or other physical conditions make stopping the train logistically difficult. One would hope that the mandatory directive would not impact the train operation immediately before the one-person crew

could safely stop the train to receive the transmission.

The different ways a multiple person crew can handle a radio communication failure also is indicative of how an FRA regulation was written with the expectation that there would be more than one train crewmember. Under most circumstances, FRA's railroad communication regulation requires a train to have a working radio in each occupied controlling locomotive, and in a second locomotive for purposes of “communication redundancy.” 49 CFR 220.9. If the controlling locomotive's radio fails en route, the crewmembers have the back-up radio in the second locomotive to use to avoid a radio blackout.

Trains with multiple crewmembers have an option not available to one-person crews. In cases of radio malfunction, it may be necessary to have a crewmember located in the second locomotive to monitor the dispatcher's communications as long as the crewmembers can otherwise communicate while the train is moving. However, if the train was a one-person operation, the lone crewmember would certainly not be able to operate from a locomotive not on the leading end, so the one-person crew would have to either try and swap out the locomotives so that the one on the leading end had a working radio to communicate with the dispatcher, or the one-person crew would need to find a way to notify the dispatcher as soon as practicable that radio communication has been lost. 49 CFR 220.38. With a multiple person operation, swapping the locomotives would likely involve a crewmember getting off the train and lining switches. Swapping the locomotives could be logistically difficult for a one-person crew depending on the track configurations encountered and the method of operation. Although a one-person crew could operate the train without a working radio to the nearest forward point where the radio can be repaired or replaced, doing so is not as safe an option as utilizing the redundant communication in the second locomotive with a working radio—an option more likely to be utilized with a multiple-person train crew.

6. Adding a Potential Safety Hazard to Highway-Rail Grade Crossing Activation Failures

The general public is directly impacted when a highway-rail grade crossing fails to activate because that means motor vehicle traffic would not receive any warning of an approaching train. Protecting the public is paramount to train operation, and FRA requires that a train can only proceed through the crossing when other steps are taken to protect highway users from approaching trains. 49 CFR 234.105. If a railroad has enough time to arrange for an equipped flagger or a uniformed law enforcement officer to be at the crossing, then the train may proceed through the crossing without stopping, albeit at potentially a slower than normal speed depending on the number of flaggers/officers. However, if a railroad does not have enough time to make other arrangements, the only other method that will allow the train to proceed through the crossing is if the train stops prior to entering the crossing in order to permit a crewmember to dismount to flag highway traffic to a stop. The flagging crewmember is not allowed to reboard the train until the locomotive has completed its procession through the crossing. Hence, under FRA's regulations, a one-person crew could not stop and flag the crossing without a non-crewmember flagger or a uniformed law enforcement officer's assistance.

Certainly, a railroad's on-time efficiency would be negatively impacted by the activation failure because a train with a one-person crew would have no choice but to wait until a flagger or officer arrived before proceeding through the crossing. Depending on the circumstances, the general public might also be negatively impacted. For example, if the train was forced to stop in a highly populated area, nearby citizens and businesses might be inconvenienced by the locomotive engine noise or exhaust fumes. Another concern is whether the train stopped clear of all other crossings. Highway users and local emergency responders may be significantly inconvenienced if the railroad and one-person train crew were unable to plan a safe place to stop the train without blocking other grade crossings. Planning a safe place to stop the train is typically considered a conductor's job, but with only one crewmember the one-person crew has no one else to help. Motor vehicle drivers or local emergency responders would not be given any advance warning of the blocked crossing or any information regarding when the crossing would no longer be blocked. Such poor planning can infuriate motor vehicle drivers and lead these drivers to take risks not to get caught waiting for a train the next time they see a grade crossing warning system begin to activate. In some cases, such poor planning could compromise the ability of local emergency services to respond. Thus, there is the potential for immediate and future repercussions when there is only a one-person train crew and no ability to quickly flag the crossing.

E. Defining the Crewmembers' Qualifications

In this proposed rule, FRA chose not to define the duties of the two mandatory crewmembers. FRA previously fulfilled its statutory obligations to promulgate regulations requiring certain minimum standards for locomotive engineers and conductors. 49 U.S.C. 20135 and 20163 and 49 CFR parts 240 and 242. FRA believes that each locomotive or train must have a crew that can perform all of the duties described by the qualifications requirements in the certification regulations for these two operating crewmembers. This can be accomplished with the assistance of technology and sometimes with the assistance of one or more other safety-related railroad employees who are not recognized by the railroad as the train's conductor. In this background, FRA will reiterate the regulatory requirements, focusing on the existing limitations and acknowledging FRA's policy. This issue is raised because FRA may consider adding requirements in the final rule specifying minimum requirements for a second crewmember's qualifications, in the event that person is not a qualified conductor. There is a question of whether the rule might need to define the duties of a freight train second crewmember who is not a conductor differently from the duties of a passenger train second crewmember.

Nearly every movement of a locomotive, whether or not the locomotive is coupled to other rolling equipment, requires that the operation be performed by a certified locomotive engineer. 49 CFR 240.7 (defining “locomotive engineer” and allowing exceptions for movements of locomotives: (1) Within a locomotive repair or servicing area and (2) of less than 100 feet for inspection or maintenance purposes). Until technology is developed that might allow for the safe operation of locomotives or trains completely by computer automation, a person is needed to operate the locomotive or train, and that person is required to be certified pursuant to FRA's locomotive engineer regulation. The issue of whether a one-person crew can operate safely is mainly an expansion of the role of a locomotive engineer to include some or all of the duties of a conductor, sometimes with the assistance of technology and sometimes with the assistance of one or more other safety-related railroad employees who are not

recognized by the railroad as the train's conductor.

In the conductor certification final rulemaking, FRA recognized that there may be circumstances where a person is “serving as both the conductor and the engineer.” 76 FR 69802, 69809, Nov. 9, 2011 (explaining that a person may hold both a locomotive engineer certification and a conductor certification, and establishing rules for when revocation of each certification is appropriate under 49 CFR 242.213). In doing so, FRA recognized the realities of remotely controlled locomotive and train operations which often involve yard or yard-type operations, travel to and from yards, or travel to service customers, without a second crewmember being present.

See

49 CFR 242.213(d). Similarly, FRA permits a certified conductor attached to a train crew in a manner similar to that of an independent assignment when a certified conductor is not accompanying a locomotive engineer or the engineer is not dual conductor/engineer certified. However, FRA expressly noted in the NPRM that the “conductor certification regulation, including section 242.213, be neutral on the crew consist issue [and that] [n]othing in . . . part 242 should be read as FRA's endorsement of any particular crew consist arrangement.” 76 FR 69166, 69179, Nov. 10, 2010. This disclaimer was made to facilitate industry-wide discussion on the conductor certification rulemaking and foster a potential consensus recommendation from FRA's Federal advisory committee, without the conductor rule becoming a referendum on the issue of crew size. Thus, although portions of the conductor rule could be read to suggest FRA acceptance of a variety of one-person crew operations, FRA's explicit disclaimer shows that the agency did not intend for the conductor rule to be that sort of proclamation.

FRA's foremost concern is that a passenger railroad will have one person in the crew who is dual certified as both a locomotive engineer and a conductor, but a second person may be lacking many of the relevant qualifications normally associated with a passenger conductor. If a second passenger train crewmember lacks too many of the qualifications of a conductor, the second person may not be truly helpful in emergency situations or even routine rail operations. The potential for creating foreseen and unforeseen problems with using a second passenger crewmember who is not conductor qualified is disconcerting. For these reasons, FRA encourages interested parties to comment on whether FRA should address this issue in the final rule. For example, FRA suggests that a second passenger crewmember who is not a conductor should be qualified on: (1) The signals to be encountered, including the name and possible indications; (2) the physical characteristics of the territory to be operated over; (3) flagging; (4) railroad operating rules (49 CFR part 218); (5) railroad radio and communications rules (49 CFR part 220); (6) passenger equipment safety standards (49 CFR part 238); and, (7) passenger train emergency preparedness (49 CFR part 239). Currently, FRA has enforced a safe course through the approval process requirement in the passenger train emergency preparedness rule. 49 CFR 239.201. Although FRA may continue to use the emergency preparedness approval process in this manner, the passenger railroad industry or public might benefit from a clear set of requirements for the qualification of a second train crewmember.

FRA has similar concerns about a second freight train crewmember who is not a certified conductor. A railroad might employ a brakeman or other operating crewmember who lacks the versatility of a conductor, which could raise questions regarding the safety of such a two-person operation. Similar operational questions could arise with the use of a person who is more like a utility employee (

see

49 CFR 218.22) than a crewmember who is assigned to a train. There are certainly some duties that a utility employee can perform for a train crew that would typically be classified as the responsibility of a freight conductor. However, because the utility employee is neither in the locomotive cab with the locomotive engineer or in near constant radio communication with the locomotive engineer while the train is moving, the utility employee cannot be deemed a replacement for all of the conductor's duties and benefits. In order to address safety concerns with the use of a second crewmember who is not a certified conductor, FRA seeks comments on whether the final rule should identify specific minimum qualifications for freight train crewmembers that lack all of the qualifications of a conductor. Minimum requirements for a second freight train crewmember who is not a certified conductor might include: (1) Knowledge of railroad rules and safety instructions; (2) railroad operating rules particular to handling equipment, switches, and fixed derails (49 CFR part 218, subpart F); (3) railroad radio and communications rules (49 CFR part 220); and, (4) brake system safety for freight trains and equipment, including end-of-train devices (49 CFR part 232).

FRA requests public comment on how railroad operations can and do safely and efficiently comply with these regulations with one-person crews or autonomous trains. Are there particular operational contexts in which compliance using one-person crews is particularly difficult or poses greater safety risks? What risk mitigating measures will railroads use to safely and efficiently comply with these regulations using one-person crews? Should any of these regulations be revised to allow one-person crews to operate safely and efficiently?

III. RSAC Overview

In March 1996, FRA established the Railroad Safety Advisory Committee (RSAC), which provides a forum for collaborative rulemaking and program development. RSAC includes representatives from all of the agency's major stakeholder groups, including railroads, labor organizations, suppliers and manufacturers, and other interested parties. A list of RSAC members follows:

American Association of Private Railroad Car Owners (AARPCO);

American Association of State Highway & Transportation Officials (AASHTO);

American Chemistry Council;

American Petroleum Institute;

American Public Transportation Association (APTA);

American Short Line and Regional Railroad Association (ASLRRA);

American Train Dispatchers Association (ATDA);

Association of American Railroads (AAR);

Association of State Rail Safety Managers (ASRSM);

Association of Tourist Railroads and Railway Museums (ATRRM);

Brotherhood of Locomotive Engineers and Trainmen (BLET);

Brotherhood of Maintenance of Way Employes Division (BMWED);

Brotherhood of Railroad Signalmen (BRS);

Chlorine Institute;

Federal Transit Administration (FTA); *

Fertilizer Institute;

Institute of Makers of Explosives;

International Association of Machinists and Aerospace Workers;

International Brotherhood of Electrical Workers (IBEW);

Labor Council for Latin American Advancement (LCLAA); *

League of Railway Industry Women; *

National Association of Railroad Passengers (NARP);

National Association of Railway Business Women; *

National Conference of Firemen & Oilers;

National Railroad Construction and Maintenance Association (NRC);

National Railroad Passenger Corporation (Amtrak);

National Transportation Safety Board (NTSB); *

Railway Passenger Car Alliance (RPCA)

Railway Supply Institute (RSI);

Safe Travel America (STA);

Secretaria de Comunicaciones y Transporte; *

SMART Transportation Division (SMART TD)

Transport Canada; *

Transport Workers Union of America (TWU);

Transportation Communications International Union/Brotherhood of

Railway Carmen (TCIU/BRC);

Transportation Security Administration (TSA).

* Indicates associate, non-voting membership.

When appropriate, FRA assigns a task to RSAC, and after consideration and debate, RSAC may accept or reject the task. If accepted, RSAC establishes a working group that possesses the appropriate expertise and representation of interests to develop recommendations to FRA for action on the task. These recommendations are developed by consensus. The working group may establish one or more task forces or other subgroups to develop facts and options on a particular aspect of a given task. The task force, or other subgroup, reports to the working group. If a working group comes to consensus on recommendations for action, the package is presented to RSAC for a vote. If the proposal is accepted by a simple majority of RSAC, the proposal is formally recommended to FRA. FRA then determines what action to take on the recommendation. Because FRA staff play an active role at the working group level in discussing the issues and options and in drafting the language of the consensus proposal, and because the RSAC recommendation constitutes the consensus of some of the industry's leading experts on a given subject, FRA is often favorably inclined toward the RSAC recommendation. However, FRA is in no way bound to follow the recommendation and the agency exercises its independent judgment on whether the recommended rule achieves the agency's regulatory goals, is soundly supported, and is in accordance with applicable policy and legal requirements. Often, FRA varies in some respects from the RSAC recommendation in developing the actual regulatory proposal or final rule. Any such variations would be noted and explained in the rulemaking document issued by FRA. If the working group or RSAC is unable to reach consensus on recommendations for action, FRA resolves the issue(s) through traditional rulemaking proceedings or other action.

IV. No Recommendation From the RSAC Working Group

On August 29, 2013, the RSAC accepted a task (No. 13-05) entitled “Appropriate Train Crew Size.” The statement clarified that “[i]n light of the recent Canadian train incident and the subsequent emergency directive issued by Transport Canada, FRA believes it is appropriate to review whether train crew staffing practices affect railroad safety.” FRA identified four purposes of this task, which were all variations on requests for RSAC to evaluate whether and how crew redundancy affects railroad safety and when crew redundancy should be deemed necessary. Crew redundancy is the idea that a second crewmember can confirm for the locomotive engineer important information thereby providing a second layer of assurance that the train is being operated in accordance with all applicable rules, procedures, practices, restrictions, and signal indications. However, the second crewmember's responsibilities are not just passive in a confirming way. The second crewmember can provide redundancy by taking the lead on tasks that free the locomotive engineer to focus on the engineer's core role of train handling.

The task statement specified that RSAC was expected to look at a list of FRA rail safety regulations to evaluate whether and how crew size impacts rail safety. The statement also asked RSAC to review published studies and reports, as appropriate. FRA provided the five FRA-sponsored studies, as well as the one TRB conference report, each of which were described previously in this preamble. In reviewing these materials, FRA was hoping that RSAC would be able to address the following issues in its recommendations report:

• Report on whether there is a safety benefit or detriment from crew redundancy, including an analysis of observed safety data and outcomes from current crew deployment practices.

• Review existing regulations and consider the impact of crew size on the performance of any task or activity.

• Report on the costs and benefits associated with crew redundancy.

• If appropriate, develop recommended regulatory language or guidance documents regarding crew size requirements that enhance the safety of railroad operations by providing enhanced regulatory redundancy. In considering the development of regulatory language, specifically consider the value of regulatory redundancy in terms of crew size as it relates to trains or vehicles identified by the group responsible for Task Number 13-02 (

i.e.,

an RSAC task to identify types and quantities of hazardous materials for special handling as a result of reviewing the Lac-Mégantic accident) as requiring special handling and/or operational controls, and if appropriate develop recommended regulatory language specific to these railroad operations.

Furthermore, in order to accommodate some RSAC members, RSAC agreed to consider other issues that have some arguable connection to the crew size issue. These other issues were to consider (1) the appropriate role and impact of technological advances on crew size and crew deployment and incorporate these into any recommendation developed, (2) PTC and Remote Control Operations or other operations where crew deployment practices or the use of technology may enhance the safety of operations, and (3) the application of a System Safety Program to these issues.

In addition to FRA, the following organizations contributed members:

APTA, including members Capital Metropolitan Transportation Authority (CMTA), Keolis North America, Long Island Rail Road (LIRR), Massachusetts Bay Commuter Railroad Company (MBCR), Metro-North Railroad (MNCW), North County Transit District (NCTD), Regional Transportation District (RTD), and San Joaquin Regional Rail Commission;

• ASLRRA, including members from Central California Traction Company (CCT), Farmrail System (FMRC), Genesee & Wyoming Inc. (GNWR), Indiana Rail Road Company (INRD), OmniTRAX, Pinsly Railroad Company, and WATCO Companies, Inc. (WATCO);

• ASRSM, including members from the California Public Utilities Commission (CPUC);

• ATDA;

• ATRRM

• BLET;

• BMWED;

• BRS;

• NRC, including members from Herzog Transit Services (Herzog);

• SMART TD;

• TCIU/BRC; and

• TWU.

The Working Group convened five times on the following dates in Washington, DC. Minutes of each of these meetings are part of the docket in this proceeding and are available for public inspection.

• October 29, 2013

• December 18, 2013

• January 29, 2014

• March 5, 2014

• March 31, 2014

As the Working Group meeting notes in the docket reflect, FRA started the first meeting by providing an overview

of FRA's position on the crew size issue. Although FRA always enters any RSAC discussion with an agency position on the issue being discussed, FRA was quicker than in previous RSAC discussions to reveal its broad-based positions. Typically, FRA will start the first meeting with a free-form discussion of the topic, allowing the RSAC Working Group's members to brainstorm problems and a range of acceptable solutions. The typical approach works well when FRA is unsure of whether a regulation is necessary, there already is an informal consensus that action needs to be taken, or the Working Group knows FRA will regulate the issue because there is a statute mandating promulgation of a regulation. None of these scenarios were present with the crew size issue. For these reasons, FRA believed it needed to approach this RSAC differently by defining its broad position on appropriate train crew size at the beginning of the first meeting.

During that first RSAC Working Group meeting, FRA presented some background on the crew size issue. FRA acknowledged that it had not previously felt the need to talk about crew size until recently for several reasons. Historically, crew size has been an issue for labor relations, and technology has enabled a gradual reduction in the number of train crewmembers from about five in the 1960s to two in 2014. Four major technological breakthroughs were mentioned in FRA's presentation that led to the historic train crew size reductions: (1) The phase out of steam locomotives allowed locomotives to be operated without crew known as fireman dedicated to keeping the engine fed with coal, (2) the introduction of portable radios made it easier to transmit information from a crewmember at the far end of the train to the leading end, (3) the end-of-train device replaced the need for one or more crewmembers to be at the rear of a train on a caboose to monitor brake pipe pressure, and (4) the development of improved train control devices helped automate safer operations in case of human error. Furthermore, FRA raised another significant technological innovation that has become widespread over the last 20 years; that is, remotely controlled locomotive operations utilizing only a one-person crew for switching service have become commonplace.

FRA told the Working Group that the agency's position on appropriate crew size is that: (1) Railroad safety is enhanced through the use of multiple crewmembers, (2) it is difficult to comply with current safety regulations and operating rules when operating with a one-person crew, (3) FRA's safety regulations were written with at least a two-person crew in mind and that operating with a one-person crew may, in some cases, compromise railroad and public safety, and (4) a second crewmember provides safety redundancy and provides a method of checks and balances on train operations. For all these reasons, FRA took the position that it needs to have some oversight of train crew size so that it can protect railroad employees and the general public.

FRA then explained its broad position on establishing train crew size requirements, explaining that the agency wanted the Working Group to make recommendations that would establish safe practices for both two-person train operations and those with less than two-persons. For instance, FRA took the negotiating position that the Working Group should develop a recommendation with a baseline of a minimum two-person crew for freight and passenger trains. The Working Group was told that FRA wanted to hear about current one-person crew operations that have been safely conducted so that those exceptions to a two-person standard could be carved out in the RSAC's recommendations. FRA also expressed an interest in offering to provide for a special approval process in a crew size regulation that would allow FRA to quickly and efficiently provide review and approval of any train crew arrangement that could not meet any easy to define specific exclusions. In order to ensure reasonable oversight, FRA suggested that a special approval would be granted based on whether the railroad's petition demonstrated an appropriate level of safety based on a combination of safeguards offered by shoring up operating procedures and implementing proven technologies. FRA noted that this was a generous compromise position, as FRA was not taking an absolute position that all trains must be operated with a two-person crew because it has the expertise to recognize accepted safe practices.

FRA's broadly stated negotiating position at the Working Group meetings was also constructed based on feedback recently received from two railroad associations participating as RSAC members. In response to Emergency Order 28, which was issued after the Lac-Mégantic accident, AAR reported to FRA that “Class I railroads currently use two-person crews for over-the-road mainline operations.”

3

AAR was certainly looking to assure FRA that the major railroads were not conducting one-person trains transporting the types and quantities of hazardous materials specified in appendix A of Emergency Order 28. ASLRRA could not be specific about each of its members' policies on transporting hazardous materials with one-person crews. However, ASLRRA tried to assure FRA that its members had “carefully consider[ed] the appropriate train and engine crew assignments to assure the highest degree of safety for the movements they operate.”

4

Taking the AAR and ASLRRA's comments at face value, FRA did not believe the agency's initial negotiating position differed greatly from the status quo. That is, the major railroads were already using two-person train crews for over-the-road mainline operations and the shortlines were carefully considering safety, presumably through a safety analysis of each operation prior to implementation—or so that was intimated.

3

Letter from Mr. Edward R. Hamberger, President and CEO of AAR, to Mr. Joseph C. Szabo, FRA Administrator (Oct. 16, 2013), which has been placed in the docket to this rulemaking.

4

Letter from Mr. Richard F. Timmons, President of ASLRRA, to Mr. Joseph C. Szabo, FRA Administrator (Oct. 17, 2013), which has been placed in the docket to this rulemaking.

Despite the AAR and ASLRRA's publicly stated positions on crew size, it was clear from the first meeting that the members of these associations were opposed to RSAC making any recommendation that provided FRA with oversight on crew size issues. AAR stated at that first meeting that there is no safety justification for FRA to address train crew size. ASLRRA took the position that because there have been very few, if any, accidents involving a one-person crew, and management has been very responsible regarding crew size, that FRA should not dictate safety regulations on the subject. FRA interpreted that unwillingness as an indication that the industry does not intend to maintain the status quo. Thus, FRA believes it cannot rely on the assurances made in the associations' written pronouncements.

As more Working Group meetings were held, FRA became increasingly concerned about the extent of one-person train operations in the U.S. and the extent that these operations may have proliferated without FRA oversight of them. Based on discussions with the railroad members of the Working Group, there appears to be a trend that more railroads of every class are willing to experiment with one-person train crew operations. Members representing Labor

organizations seemed as surprised as FRA with some of the generalized statements made by a variety of railroads regarding the extent of the existing one-person operations. For example, railroads of all classes seemingly have permitted remote control operations with only one-person to routinely operate on main track in limited train service, as opposed to being used for switching service—the original expected use for which the technology was designed. AAR and ASLRRA were unwilling to recommend FRA oversight of their members to assure railroad employees and the general public that their members' existing operations are safe, proclaiming that the lack of safety data showing there was an existing problem should prevail as an argument.

Without a requirement for railroads to consult FRA on questionable crew size practices, FRA did not field inquiries from railroads asking for the agency's opinion on the safety of the practices. Even if an FRA inspector were to observe a train being operated with only one-person, FRA personnel would not have any reason to write up an inspection report detailing the finding—unless the one-person operation was alleged to have violated an FRA safety law, regulation, or order and the issue was tangentially raised in the report. Certainly, high level safety personnel at FRA were unaware of how many railroads, especially freight railroads, were regularly fielding trains with only a one-person crew. For these reasons, the Working Group's discussions of existing one-person train crew operations were illuminating.

Just as railroads have explained for over a century that certain operating rules were “written in blood” because it took one or more accidents causing serious injuries or fatalities before the operating rule was written, railroad employees and the general public should not have to wait for horrific accidents before the Federal government takes action. FRA provided the Working Group with a number of significant reasons for recommending regulatory action. In summary, FRA provided: (1) The scientific research studies showing the benefits of a second crewmember, (2) the anecdotal information regarding recent train accidents and how a second crewmember either could have played a safety role or did play such a role, (3) the explanation that FRA's railroad safety regulations were written with the expectation that nearly every train would be operated by no fewer than two crewmembers, and (4) the general public's negative reaction to the idea that FRA did not already mandate two-person train crews to add another layer of safety.

During the Working Group's first meeting, SMART-TD stated its belief that FRA appears to be responding to the public's demand for action. SMART-TD backed up its statement during the Working Group's January 29, 2014, meeting when it shared a research report it sponsored that combined data from five surveys that indicated a strong level of bipartisan support among voters for a Federal law requiring freight trains to operate with a crew of two. The surveys were conducted in the States of Kentucky and North Dakota, and in select Congressional districts in the States of Colorado, Kansas, Iowa, and Pennsylvania. The data supported a finding that 77 percent of all respondents support Federal legislation requiring freight trains to be operated by a crew of two. Even when respondents were not reminded in a prior question about recent deadly train accidents in Quebec, Spain, and New York City, 74 percent supported Federal legislation. Another finding was that an overwhelming majority of those polled (between 83 to 87 percent in each of the five surveys) had the opinion that, generally speaking, when it comes to railroad safety and operations, one operator cannot be as safe as a train with a crew of two individuals. A copy of this report has been placed in the docket.

Despite the early warning signs that the Working Group would not be able to reach a consensus, FRA held 5 day-long meetings spread out over 6 months in which the agency continued to make substantive presentations and negotiate in good faith. Every time APTA or ASLRRA presented a new set of facts for a potential exception, FRA listened and came back with a written recommendation that tried to capture the request for leniency. Twice, AAR provided the Working Group with a list of a variety of railroad operations that it claimed should be allowed to continue with one-person with no restrictions. Each time, FRA responded with a written recommendation that tried to capture the request for leniency or, in a few instances, explained why it could not support such a request. Although no consensus was reached during the Working Group meetings, there seemed to be a tacit understanding that FRA had adequately described each operation for which it included an exception in its working document.

First, at the January 29, 2014 meeting, AAR listed the following examples as non-revenue movements that it suggested should not require a minimum of two crewmembers: “(1) Helpers; (2) Pushers; (3) Light engines; (4) Passenger moves; (5) Hostlers; (6) Locomotive exchange crews; (7) Work trains; (8) Wreck crews; and (9) Roadway maintenance machines.” Final Minutes 2014 0129 TCWG-14-03-0503 pdf at 15. During the same meeting, AAR also asked whether FRA would agree to an exception for (10) interchange and transfer moves, (11) mine load out or plant dumping, and (12) toxic by inhalation or poisonous by inhalation (TIH/PIH) hand-offs, where one crewmember remains behind to facilitate secure hand-off, a Transportation Security Administration (TSA) requirement. FRA agreed, and altered its Working Group proposal to include an exception for each of the twelve items with the following caveats: (1) FRA did not believe a special exception was necessary for pushers, as the exception for helpers also covers pushers; (2) FRA provided an exception for light/lite engines, but made clear that the exception did not apply to passenger diesel or electric multiple unit (DMU or EMU) operations; (3) FRA provided an exception for hostlers conducting switching operations, but not hostlers working in other than switching operations; (4) FRA considers a wreck crew to be a work train, and FRA provided an exception for work trains; (5) FRA's work train exception applies to roadway maintenance machines in a work train, but such machines are not otherwise excepted; (6) FRA did not except interchange/transfer train movements as these operations, which may travel up to 20 miles while picking up or delivering freight equipment under the definition of “transfer train” in 49 CFR 232.5, pose the same safety issues as other trains that are not limited to traveling 20 miles; and (7) during a TIH/PIH hand-off, FRA did not create an exception that would allow the second crewmember to be left behind with the PIH/TIH car while the train departed with only a one-person crew as the train continuing would pose the same safety issues as other trains.

Second, in anticipation of the final Working Group meeting held on March 31, 2014, AAR submitted a document on March 28, 2014, titled “Discussion of Current Class I Operations Using Vehicles When Assisting Trains.” AAR Discussion Document TCWG-14-03-31-04.pdf. The document describes six situations where a second train crewmember would need to be located outside of the operating cab of the controlling locomotive when the train is moving in order to continue to perform the duties assigned, and then lists seven

additional examples. The second train crewmember would then need another way to catch up to the train to get back on it. FRA believes all of the operations described in that AAR document are acceptable, as long as the second train crewmember that is separated from the train can directly communicate with the crewmember in the cab of the controlling locomotive pursuant to proposed 49 CFR 218.125(d). FRA has greatly benefited from the open, informed exchange of information during the meetings. Although the Working Group did not reach consensus on any recommendations, FRA decided not to extend the April 1, 2014, deadline that FRA initially presented the RSAC. FRA did not think it would be beneficial to continue to discuss with the RSAC's railroad members the issue of what data FRA had to support this rulemaking recommendation when they knew full well that the data, supplied by the railroads themselves to FRA, does not capture accidents where the cause or contributing factor was lack of a second crewmember.

It was also made clear to FRA that organizations representing railroad employees supported FRA's overall concept of mandating two-person crews on each train with some exceptions, but were overwhelmingly opposed to FRA's draft rulemaking recommendation that attempted to greatly accommodate all classes of passenger and freight railroads. Several labor organizations wanted FRA to scale back some of the exceptions FRA accepted as part of the agency's attempt to reach a consensus. For example, these organizations wanted to limit the shortline railroad exceptions in 49 CFR 218.131(a) to a freight train operated on a railroad and by an employee of a railroad with 15 or fewer employees, rather than the FRA position of “a freight train operated on a railroad and by an employee of a railroad with less than 400,000 total employee work hours annually” (which is the equivalent of about 200 or fewer employees). Labor organizations also expressed a preference for requiring each railroad to petition for a waiver to utilize less than two train crewmembers rather than recommend a special approval procedure that would propose a much shorter FRA review period. Thus, after five meetings, with labor and management representatives taking polar opposite positions on large and small issues, FRA decided not to accept some Working Group members' recommendation to extend the deadline for negotiating a recommendation.

V. FRA's Overall Post-RSAC Approach

This proposed rule offers a pragmatic approach to providing oversight of the crew size of non-switching train services to ensure the continued safety of railroad employees and the general public. In that respect, FRA's approach to the crew size issue has remained the same as when the agency first brought its position to the Working Group's attention. FRA views its crew size concerns as a relatively small current problem that has the potential to balloon into a much greater problem in the not-too-distant future if appropriate oversight is not exercised. Because there is significant potential for this safety issue to become a much greater problem in the second half of this decade, FRA believes the time to act is now.

A. The Proposal Is Largely Focused on Influencing How Railroads Approach Future One-Person Operations

Based on information orally provided by AAR regarding the major railroads current train crew size practices, it appears that the proposed rule would not have a substantial impact on the current operation of the major railroads. Each major railroad appears more concerned about how a crew size regulation would impact the railroad's possib

This text is long and has been trimmed here. Open the source document for the complete record.