This paper presents an interactive, fault model-based prototype diagnostic tool that will assist service technicians in isolating the root cause of vehicle problems and performing corrective repairs. Current automotive service procedures are driven primarily by static service manuals that inform technicians on the service steps in case a specific diagnostic trouble code (DTC) is set in a vehicle. Although comprehensive, these service procedures usually require technicians to gather and integrate diagnostic information from several sources, such as DTCs, customer complaints and manual test results. This can lead to increased repair time and labor costs.
The fault model-based interactive service procedure tool discussed in this paper will guide the technician to isolate the fault and provide him/her with recommendations for the correct repair actions. The tool uses a fault model, built using service procedures information, historical repair data and engineering inputs. This fault model captures dependencies between multiple symptoms such as DTCs, vehicle operational parameters, customer complaints, and technician test outcomes and relates them to probable faults. In the service bay, this tool will employ the fault model and vehicle symptoms to provide a ranked list of suspect components, related repairs and the next steps necessary to further isolate the fault. In effect, as the technician collects and inputs more symptoms or test outcomes into the tool, it will further isolate the fault to specific components and recommend further actions till the fault is isolated to one specific component; and then recommend the correct repair. The capabilities of this system are demonstrated on an automotive fuel evaporative emission control system.
This tool will enable correct and efficient first-time repairs leading to a reduction in warranty costs and repeat visits.