This paper describes the design of an expert system used as a part of a
diagnostic system for diesel hydraulic drives. A PC-based diagnostic system ruDi (KeBler and Burgwinkel, 1993)
was developed to minimise the time for diagnosis and repair work of
vehicles with diesel hydraulic drives used in the mining industry. One part
of ruDi is an expert system that helps to locate errors in any of a vehicle's
subsystems, and supports technicians in removing the cause of a
malfunction. Apart from the standard knowledge of diesel engines and
hydraulic transmissions, the expert system supports the interpretation of
the data obtained during sophisticated engine analysis tests such as
high-resolution speed analysis or high-pressure analysis of the diesel
injection system. The expert system is designed to diagnose various vehicles of the same
type. Data for each vehicle is stored separately allowing checks to be
carried out individually for each machine. An additional feature of this
system is a database interface to a Maintenance Planning and Control
System (MPCS) that allows an automatic generation of work orders for
the defects found during the expert system sessions.