Applications of Expert Systems in Transport

BACKGROUND

Experienced judgement and specialist knowledge are essential to the proper specification, understanding and interpretation of data and computer analyses. The human expert has traditionally supplied this knowledge and judgement with the computer doing the necessary number-crunching. However, artificial intelligence (AI) research provides ways of embodying this knowledge and judgement within computer programs. Despite an early lead in the field, UK research and developmnent into AI techniques was held back in the 1970s when the then Science Research Council took the view that the 'combinatorial explosion' of possibilities would be an insurmountable obstacle to AI developent. But in America and Japan research continued, and the surge of interest in the 1980s has been a consequence of the 'Fifth Generation Computer' research programme initiated by Japan (Feigenbaum and McCorduck; 1984). This led in Europe to the ESPRIT programme of advanced technology research, and in the UK to the Alvey programme (Department of Industry, 1982). As a result, all sectors of industry have been encouraged to consider how such advanced technology can be applied, and the transport industry is no exception.

This paper sets out to explain some of the relevant techniques in simple terms, and to describe a number of situations in which transport planning and operations might be helped through their use, illustrating this by reference to the pioneering work going on in transport applications in the USA, Britain and Australia.