Sir George Findlay was a prominent British railroad manager. He was the general manager of the London & Northwestern Railway, a major company that underwent nearly 20 years of expansion under his leadership.
His 1895 book, The Working and Management of an English Railway, described an organization for railway track maintenance. The basic gang had four people responsible for 4 miles of track. Supervision and management included inspectors, a chief inspector, and a divisional civil engineer. Each division also had draftsmen, masons, and other special crafts at their disposal.
Findlay described the maintenance organization for railway signals. He emphasized the importance of keeping signal systems in perfect working order to prevent collisions. Each signal-house was visited by the “chargeman” every two weeks for lubrication and minor repairs, and monthly by a more capable “fitter”. The chargeman would report followup work to the local inspector, and the fitter filed a report with a central office. More extensive repairs were assigned to a repair gang.
Inspectors would examine freight cars before leaving a station for good working condition, including springs, wheels, and checking for hot bearings:
“…any waggon in which a defect is discovered is promptly shunted out of the train, and a red card affixed to it to indicate that it is not to run until the necessary repairs have been attended to.”
This readiness checklist prevented existing wagon problems from becoming operational failures of the entire train. The red card was a visual defect tag, a best practice adopted by Lean practitioners.
Findlay noted that there was a system to exchange lessons learned from maintenance problems:
Periodical meetings of the divisional engineers are held, at which the various points which arise from time to time in connection with the maintenance and repair of the permanent way are discussed, and the engineers are thus enabled to compare notes and give each other the benefit of their respective experience.
This brief description shows that there was an expectation for identification and resolution of maintenance problems, and dissemination of the solution to other engineering divisions.
The Western Railway Club
The Western Railway Club is a railroad trade and professional association that has operated since 1898. In the 1900s, the club held monthly meetings in Chicago’s historic Auditorium Building Hotel.
Attendees were normally senior business and technical managers, ranging from division supervisors, to railway vice-presidents and presidents. The Western Railway Club’s monthly meetings were sometimes quite technical and often covered maintenance or reliability.
During a meeting in September 1906, the club’s Secretary, with more of a background in insurance than railroad management, introduced Chief Inspector James Foord of Hartford Steam Boiler Inspection and Insurance Company. Foord presented a summary of Hartford’s boiler inspection program. Foord related the Hartford company’s background and the experience of the Manchester Steam User’s Association. Both the Hartford and Manchester programs started in the 1860s and could demonstrate several decades of improvement in boiler safety and reliability through periodic inspections. The improvement of boiler reliability by preventive inspections had been debated in the English speaking world throughout the 19th century. Foord used photographic slides from explosion investigations to explain Hartford’s root cause investigation techniques and typical findings.
Dunham at Western Railway Club
Walter Dunham graduated from Cornell in 1895, and had been a draftsman and master mechanic for two major railroads. The master mechanic was the engineer who supervised maintenance of the locomotives and cars. Dunham proposed standardizing engine failure reports to allow methodical defect elimination.
Dunham’s proposal was read before the Western Railway Club on January 15, 1907. The Club secretary actually presented the paper because Dunham was not actually there. Excuses were given that he was attending to a train breakdown.
Dunham began by recognizing that operational failures can be the results of operator errors, in part due to lack of rest. He further recognized that these failures are often misattributed to the repair organization:
…an engine crew should be given as much consideration as we do our engines. They can not work perpetually on short hours of rest unless their working hours are made short.
It is natural during a rush period, whether of long or short duration, that engine failures increase in number and often also in proportion to the total miles run in a given period of time. These results may be due more to the operating methods and conditions than to any lack of attention on the part of the roundhouse forces. Under these circumstances the division as a whole should stand the results, instead of the motive power department only being charged with a failure.
Most railroads considered any late arrival to be an engine failure. A late train increased the chances of collision by disrupting the track control systems. Even if a collision was avoided, trains and cars could be stranded, and passengers and goods would be late.
Dunham wanted to be able to determine why a train was late, so that management could take appropriate action. As a starting point, management had to be able to distinguish between operational and mechanical failures. He proposed 13 types of delays that should not be considered engine failures, but instead were caused by decisions made by the train dispatcher. If an engine had to wait and ran out of water or coal while idling, this was not an engine failure. Delays to due weather, overloading, or operational deferral of repairs, were not engine failures in his system.
Dunham recommended summarizing types of failures, then having master mechanics and superintendents review them every month. The standardized failure reporting was to guide the entire organization to corrective action:
There is a natural tendency for enginemen to be rather cautious about sending telegraph reports of defects or troubles on snap judgment as well as a similar tendency for a dispatcher to be inclined to blame the engineman or the engine for being the primary cause of a train delay. These active representatives of two branches of the operating department are the men who can do the most toward reducing engine failures by keeping in close touch with each other when on duty and by giving all the facts in a plain and full manner when trouble does occur. If the failure is due to long hours on the road the information is of value to the superintendent, who can determine the necessary action for increasing the speed of the train or trains. From these same records his attention is forcibly brought to the results of inferior coal, poorly designed and operated coaling stations, scanty and bad water supplies, overloading of engines, indifferent train-dispatching, lack of harmony in action on the part of the men in charge of the trains.
Dunham recognized the classic problem of having operations and maintenance point fingers at each other. A formal failure classification system, he thought, would build a complete picture of failures, including latent organizational problems.
The club meeting minutes showed examples of pre-printed component failure reports used by Dunham’s railway, the Chicago & Northwestern. For major components, several drawings of standard parts were presented for the mechanic to mark up. The technical details of the failure could be recorded and reported using a standardized way. This is the first step in a formal failure reporting and corrective action (FRACAS) system.
The drawings appealed to people with a background in drafting and little long-hand writing was required. During the later presentation to the Master Mechanic’s Association, Dunham added to the exhibits, including a log form for engine delays and a ten-day summary sheet for engine delay time. The summary of delays included the event even if the delay was made up somehow, so near-miss events were included the Dunham’s system.
The standardized failure reporting was the first step in the formal FRACAS system. The analysis phase started with the ten-day summary sheet was the first step in a formal analysis. The next step was the monthly review by the master mechanic and superintendent, who were expected to prioritize and correct problems.
The club secretary had sent Dunham’s paper to several members before the meeting, and asked for examples of failure reports and policies from member railways. Several major railways sent their forms:
- Buffalo, Rochester, & Pittsburgh
- Central Vermont
- Chicago, Rock Island, and Pacific
- Central Railroad of New Jersey
- New York, Ontario, & Western
- Chesapeake and Ohio
- Canadian Pacific
- Chicago, Burlington, & Quincy
The Central Railroad of New Jersey, instead of a blank form, included their actual monthly failure summary for October 1906. The sample forms and other policy documents were entered into the meeting minutes and published. This allowed every recipient of the club minutes to review failure reporting forms and procedures from nine major railways. The club minutes were distributed to members, other organizations, and major libraries. This exchange of best practices meant that many large railways had access to each other’s methods and some data.
The club minutes also recorded the discussion of Dunham’s paper. Some members were enthusiastic and in complete agreement with Dunham’s approach. W. G. Wallace summarized his view by saying,
Engine failures—men failures—or failure to prevent failure—result in delays to trains and reduction in tonnage, and at times it appears that if the same effort were expended in preventing the failure as is manifested in fixing up the engine failure report after the failures have occurred, there would be quite a reduction in the number of failures and a corresponding improvement in the service.
Wallace was the author of a regular column featuring technical questions in the Locomotive Firemen’s Magazine titled “Locomotive Running and Repairs.” Later that year he published the collected wisdom in a handbook called Locomotive Breakdown Questions Answered and Illustrated.
Some members thought that failures were the engineer’s responsibility. In their opinion, no special reporting system was needed, because the engineer should discover and react to them as part of their normal responsibilities.
There was some criticism of the available failure modes. Dunham’s modes were focused on the physical failure and did not include many operator and dispatcher problems. The locomotive engine crew’s coal-stoking skills could cause several boiler problems, but so could poor quality coal, operating with a dirty firebox, and operating with clinker buildup in the flue. The mechanics had no control of coal quality or infrequent cleaning. Dunham’s forms required exact reporting on the nature of a fracture, but did not include the operating details like the weight of cargo, total run time, or the engine crew’s fatigue level that Dunham claimed was so important. The failure report focused on the locomotive, but not the organizational situation.
One member related that the drawings were an improvement, since he had trouble encouraging detailed reports in his own free-form fields. As anyone who operates a maintenance management data system knows, long text fields lack structure, grammar, uniformity, and are often blank. This problem pre-dates the computer.
Several members related that the dispatcher could manipulate the report data. There were several anecdotes about repair and operating departments squabbling over who had to report the events. The existing reporting policies and KPIs encouraged bureaucratic finger-pointing.
M. K. Barnum was the Western Railway Club’s new vice president and a corporate officer of the Chicago, Burlington, & Quincy Railway. During the discussion, Barnum stated the goal explicitly, to accumulate a failure database that could be analyzed for improvements to eventually reduce the failure rate:
I think that one of the most important suggestions in the paper is that pertaining to reports of defective machinery which are sent to the Assistant Superintendent of Motive Power and through him to the Mechanical Engineer for his files. By faithfully following up such reports a very valuable record will accumulate which will enable the weak points of the various classes of locomotives to be strengthened and thereby materially decrease the engine failures. (p. 186)
The club president ended discussion by recommending that standardized failure reporting be studied by the Master Mechanic’s Association.
Dunham at American Railway Master Mechanics’ Association
The 40th convention of the American Railway Master Mechanics’ Association was held six months later in Atlantic City. Dunham was not present at this convention either, and his paper was instead read by another Chicago & Northwestern master mechanic, E. W. Pratt.
The text of the paper “Locomotive Failures, Records, And Results Of Keeping Them” was nearly identical to the text recorded in the Western Railway Club minutes. The Master Mechanic’s Association minutes included Dunham’s own Chicago & Northwestern failure report forms, but did not include the exhibits from the other members of the Western Railway Club. Discussion by the master mechanics ran four pages, covering similar themes as the discussion in Chicago. Repeated discussion points included operator training and apprenticeships, skillful use of fuel, and the role of the train dispatcher.
E.A. Miller of the New York, Chicago, and St. Louis Railroad (the “Nickel Plate Road”) related that a similar failure reporting system had been in place and had reduced operational delays over the course of two years. Between his system and the Western Railway Club members who submitted exhibits, there were ten railways with documented administrative systems for failure reporting and defect elimination in 1907.
A representative from the Central of Georgia Railway proposed aligning the definition of an engineer failure with the American Railway Association. A subcommittee decided to assign a report on standardized definition of engine failure for the next year’s convention to M. K. Barnum, the vice-president of the Western Railway Club who had also been at the original reading of the paper in Chicago.
Dunham in Locomotive Firemen & Enginemen’s Magazine
The Brotherhood of Locomotive Firemen and Enginemen’s Magazine reprinted Dunham’s paper and the entire minutes of the Western Railway Club’s in the September and October 1907 editions. They did not reprint the exhibits.
The President of the Western Railway Club had closed the January discussion with the comment,
This is one of the most interesting meetings we have had for a long time and I hope we will continue to have meetings of this character. I note with a great deal of pleasure the improved attendance at the meetings, and I feel sure that the interesting subjects we have had have been responsible to a great extent for the improvement in attendance.
In the October reprinting, Mr. Dunham responded,
The true function of any system of failure reports is to give absolute facts upon which can be based the actions taken to avoid a repetition of the same or similar failures.
Dunham’s response only appears in the Locomotive Firemen’s and Enginemen’s Magazine version. It must have been included after the fact as an allowance to the author, since the comment was inserted after the Western Railway Club president moved to close the discussion.
The Locomotive Firemen & Enginemen’s Magazine also reprinted a lengthy discussion of boiler explosions, prevention of explosions, and boiler inspection systems from “The Railroad Gazette” (pg. 324). This showed that there was interest in safety, reliability, repair, and maintenance practices by the readership of the magazine, locomotive operators. Interest was not just limited to repair engineers and managers.
Failures discussed by International Railway General Foremen’s Association
Seven years later, a long discussion of failure reporting, human error, and dissemination of lessons learned from failures is recorded in the minutes of the International Railway General Foremen’s Association ninth and annual conventions in 1913 and 1914 in Chicago. The 1913 paper was titled “Engine House Efficiency” and was given by Chicago & Northwestern master mechanic Walter Smith. He detailed an engine inspection process and maintenance of shop equipment. The convention included a tour of Chicago & Northwestern’s maintenance shops. The committee that heard the paper resolved to study engine house efficiency in greater depth for the next convention. E.W. Pratt, who had read Dunham’s paper to the Master Mechanic’s Association in 1907, attended the convention, but was not a part of this discussion.
The next year, during the tenth annual convention in 1914, opening remarks touched on Scientific Management as a method to eliminate waste from all business activities. Walter Smith had extended his paper on how to run an engine repair shop efficiently. Failure reports were part of the technical system:
There are few fields of study more fruitful of results than a study of the causes of failures. Each new failure should be carefully studied, as well as the past failures shown on the records. In this way it is possible to arrive at conclusions, and thus take action to prevent re-occurrences.
Smith described some locomotive design improvements to reduce maintenance. He also described inspection procedures for locomotives after repair: “Rapid and accurate locomotive inspection is a matter of the greatest importance from a maintenance standpoint.” The procedure only allowed 4-5 minutes to ensure that the engine leaving is in good working order and ready to dispatch.
Part of the paper examined if it was better to assign a crew to the same locomotive all the time, or assign a crew to the next available locomotive, a “pool” system. One attendee reported significant reductions in repair costs and improved availability by assigning crews to a specific locomotive. He concluded that the main cause of the improvements was that the regular crew took better care of their locomotive than a rotational crew. Specifically, they followed up on correcting minor problems more than a rotational crew would. Eight other members agreed, all noting that the operations department preferred the rotational scheme, but that the maintenance and repair departments saw better performance from a policy of regular assignment. The organization president pressed the group to make a formal recommendation for all railroads to adopt this practice, but the attendees were reluctant. This is analogous to today’s discussions about operator ownership, operator basic care, and total productive maintenance.
A new discussion titled, “Prevention of Failures by Necessary Precautions” was close to Dunham’s main idea from years earlier. The representatives compared their own company definition of failure and what followup procedures existed. The representatives shared different approaches to reducing the rates of engine failures.
In 1907, Dunham represented the Chicago & Northwestern Railway at the Western Railway Club meeting and the American Railway Master Mechanics’ Association convention in 1907. The 1914 convention of the International Railway General Foreman’s Association had two vice-presidents and two committee chairmen from Chicago & Northwestern, in addition to other members. Dunham was not listed as a member or attendee for either the 1913 or 1914 foreman’s convention, but he was still with the company as Supervisor of Motive Power and Machinery. E. W. Pratt, Walter Smith, and M. K. Barnum all represented Chicago & Northwestern’s repair and maintenance group. Apparently, the Chicago & Northwestern had continued organizational interest in failure reporting and it was not the work of a single person. Dunham continued to contribute to trade journals into 1932, always listing an association with Chicago & Northwestern.
In terms of exposure, the International Railway General Foremen’s Association had 219 members and was holding their tenth annual convention in 1914. The member list shows many Chicago residents. The Master Mechanic’s Association had 819 members, and their first convention was held in 1868. By 1927, it had merged into a division of the American Railway Association, which itself joined today’s American Association of Railroads at its founding in 1934. The Western Railway Club, founded in 1898 and still operating, had 1352 members when Dunham’s paper was read at their 1907 meeting. The 1907 and 1913-1914 failure reporting discussions had a direct audience of several hundred railway engineers, managers, and vice presidents. The minutes of each convention were reprinted in other trade journals, reaching thousands. In 1907, at least ten American railroads had formal failure reports and an administrative system to act on them. The railroads were tracking downtime as small as one minute, sometimes at a subassembly level, and were pursuing industry standardization of the reporting systems.
The situation was far from settled. Five years later in 1920, International Railway General Foremen’s Association held their 15th convention. The first paper was “Standardization of Engine Failures and Terminal delays.” The foreman’s association was still working on an industry standard for how to report and classify delays, breakdowns, and failures in a way that could be adopted by the entire rail industry. Engineers from the Chicago & Northwestern were still leading the attempt at standardization.
The Working and Management of an English Railway https://hdl.handle.net/2027/coo1.ark:/13960/t3st8541d pg. 61, 145
“Engine Failures and Their Report” Western Railway Club https://hdl.handle.net/2027/wu.89064399199 pg. 151
“Locomotive Failures, Records, And Results Of Keeping Them” American Railway Master Mechanics’ Association, 40th convention, https://hdl.handle.net/2027/wu.89064399561, pg. 308
Brotherhood of Locomotive Firemen and Enginemen’s Magazine https://hdl.handle.net/2027/hvd.li2a3w, Vol 43 No 3 of Sept 1907, pg. 299, and Volume 43, Number 4, pg. 517
International Railway General Foremen’s Association 9th convention, 1913: https://hdl.handle.net/2027/uiuo.ark:/13960/t10p5983m, p. 142, 152
International Railway General Foremen’s Association 10th convention, 1914: https://hdl.handle.net/2027/uiug.30112106511386, p. 50, 59, 67, 71
Dunham in 1932: vol 90 of Railway Age in June 18, 1932
Master Mechanic’s Association merges: National Bureau of Standards 1927 Standards Yearbook, p. 324
Dunham’s position in 1913: https://hdl.handle.net/2027/uc1.b3072160 p. 368
International Railway General Foremen’s Association in 1920: https://hdl.handle.net/2027/mdp.39015039383727 Oct 1920, pg. 20, pg. 207
Barnum’s biography: https://archive.org/details/railwayagegazett88newy/page/496/mode/1up?q=barnum