Computer diagnostics aid railcar maintenance
Computer programs that mimic human problem-solving techniques offer the transit industry a potential for substantial cost savings by using advanced software technology to diagnose problems and decrease transit equipment repair time, according to the report “Artificial Intelligence for Transit Railcar Diagnostics.”
Conducted by the Transit Cooperative Research Program (TCRP) and sponsored by the Federal Transit Administration, the report states that rail car maintenance is a hot candidate for Artificial Intelligence (AI) technology, which is currently used in telecommunications, nuclear power, health care, automobile, spacecraft and electronics industries.
The technology also is used to assist technicians in the diagnosis of railcar system problems.
“Properly developed software can literally be loaded onto a laptop computer and used, in quick order, by the maintenance mechanics,” says American Public Transit Association (APTA) Vice President-Rail Transit Peter Stangl.
“For the transit customer, this means the prospect of more railcars being available for service during each rush hour.”
In the constant struggle to find innovative ways to contain costs, transit managers identify maintenance diagnosis time as a prime target for improvement.
Arinc Inc., Annapolis, Md., is one company working toward that end. Specializing in the development of traffic management and communications systems, the company is involved in systems and technologies for traffic management, traffic control, monitoring and vehicle and wayside system elements. Its packet switching network handles more than 1 million messages per day.
Acting in conjunction with the Federal Highway Administration (FHWA), the company currently is testing and evaluating communications technologies for the Intelligent Transportation Systems Program (ITS).
According to the report, AI programs such as those listed above can provide personnel with answers to problems, suggest courses of action or act automatically by being imbedded into the vehicle’s operating system. The same technology can also be used on the complex mechanical systems now being employed in the nation’s bus fleets. “The bottom line is quicker diagnosis time, faster turnaround in the shop and, ultimately, fewer maintenance costs,” Stangl says.
By using data from APTA’s 1992 Annual Financial Statistical Report, the authors of the report developed a statistical model based on a generic transit authority of 600 vehicles.
The report indicates that and AI diagnostic tool that initially costs $172,000 would have to provide only a 7.2 percent reduction in the mean time rate of repair of a railcar propulsion system to pay for itself in one year.
The report presents results of seven AI techniques: expert systems, case-based reasoning, model-based reasoning, artificial neural networks, computer vision, fuzzy logic and knowledge-based systems.
For more information on this report, contact the American Public Transit Association, 1201 New York Ave., N.W., Washington, DC 20005; (202) 898-4000.
