CBR for troubleshooting aircraft on the flightline

Knowledge-based approaches to the solution of aircraft maintenance problems are beginning to gain gradual acceptance within the airline business. The industry began several years ago by using artificial intelligence to address complex maintenance planning and scheduling problems. More recently, focus has shifted to other important areas, such as system condition monitoring and failure diagnostics. Case-based reasoning (CBR) has recently emerged as a good candidate technology to support complex system diagnostics, particularly when aiding human decision-making during time-constrained operations. However, a number of human factors and technical issues remain to be resolved if the apparent potential of CBR technology is to be both accepted and successfully exploited within the airline industry. This paper emphasises some of the findings and lessons learned during CBR based work at British Airways, and describes some of the author´s emerging views of how such future systems must be both planned and developed.<>