Download Free Artificial Intelligence For Transit Railcar Diagnostics Book in PDF and EPUB Free Download. You can read online Artificial Intelligence For Transit Railcar Diagnostics and write the review.

This report documents and presents the results of a study to determine the feasibility of applying Artificial Intelligence (AI) techniques to the diagnosis of transit railcars. The AI techniques investigated were expert systems, case-based reasoning, model-based reasoning, artificial neural networks, computer vision, fuzzy logic, and a procedural knowledge-based system. Site surveys were conducted at transit railcar maintenance facilities and at railcar subsystem suppliers. The site surveys gathered information about current and future diagnostic and maintenance practices, possible barriers to implementing advanced AI technology, and maintenance cost data. An economic analysis was performed to provide an estimate of cost savings expected by reducing the diagnostic effort.
The report documents the results of a demonstration of an AI-based program that acts as a "diagnostic assistant" for transit railcar propulsion systems. The diagnostic program uses a hybrid AI approach with both model-based reasoning and expert system rules. The AI tool was tested at the Washington Metropolitan Area Transit Authority (WMATA) on direct current chopper propulsion systems of the 3000 series railcars. The system was determined to be easy to use and effective in diagnosing propulsion system faults.

Airport Research Needs

National Research Council (U.S.). Transportation Research Board. Committee for a Study of an Airport Cooperative Research Program

Published: 2003

Total Pages: 116


Get eBook

Urges the US Congress to establish a national airport cooperative research program. The committee that produced the report called such a program essential to ensuring airport security, efficiency, safety, and environmental compatibility.