An Agent-Based Inference Engine for Efficient and Reliable Automated Car Failure Diagnosis Assistance

There are many difficulties and challenges involved in cars failure and malfunction diagnosis. The diagnosis process involves heuristic and complex series of activities and requires specific skills and expertise. A basic toolkit and assistance software are imperatives to help the car drivers to at least identify the source of car failure or malfunction, especially, when the location of the event does not permit immediate help. It enables the car driver to take an initiative in knowing the car condition and try to repeat the car. Expert systems are widely used to embody the diagnosis expertise into machines. However, improving the expert systems' inferencing capability and diagnosis accuracy are still open research topics. Consequently, this paper proposes an agent-based inference engine for the car failure diagnosis expert system that is named automated car failure diagnosis assistance (ACFDA). The agents' goal is to maximize the efficiency of the overall performance of the ACFDA system by deliberating a number of inferencing tasks and tuning the inferencing logical flow. Additionally, the agents' collective effort provides reliable solutions that best fit the users' inputs. The ACFDA system is experimentally tested by 15 relevant candidates. The test results show that the system efficiently and reliably performs the diagnosis to the most given car failure cases. The system can be integrated into cars or can be used as a separate gadget to assist the car drivers in car failure diagnosis and repair.