-
公开(公告)号:US20200184742A1
公开(公告)日:2020-06-11
申请号:US16212120
申请日:2018-12-06
Applicant: GM Global Technology Operations LLC
Inventor: Shengbing Jiang , Chaitanya Sankavaram , David A. Frank , Brian L. Buck , Janelle Kane Rolando
Abstract: A method of root cause diagnosis of fault data from a vehicle includes identifying a first vehicle fault and selecting from field repair data a vehicle feature corresponding to the identified first vehicle fault. The method also includes identifying from the field repair data an effective repair of the identified first vehicle fault. The method additionally includes training and testing via a machine learning algorithm, a labor code classifier using the identified effective repair of the first vehicle fault and the selected vehicle feature corresponding to the identified first vehicle fault. The method also includes identifying and classifying, using the trained classifier, indistinguishable labor codes. Furthermore, the method includes communicating the identified and classified indistinguishable labor codes for diagnosing a root cause of real time first vehicle fault data. A computer-readable medium storing an executable computer algorithm for performing the root cause diagnosis of vehicle fault data is also envisioned.
-
公开(公告)号:US11151808B2
公开(公告)日:2021-10-19
申请号:US16212120
申请日:2018-12-06
Applicant: GM Global Technology Operations LLC
Inventor: Shengbing Jiang , Chaitanya Sankavaram , David A. Frank , Brian L. Buck , Janelle Kane Rolando
Abstract: A method of root cause diagnosis of fault data from a vehicle includes identifying a first vehicle fault and selecting from field repair data a vehicle feature corresponding to the identified first vehicle fault. The method also includes identifying from the field repair data an effective repair of the identified first vehicle fault. The method additionally includes training and testing via a machine learning algorithm, a labor code classifier using the identified effective repair of the first vehicle fault and the selected vehicle feature corresponding to the identified first vehicle fault. The method also includes identifying and classifying, using the trained classifier, indistinguishable labor codes. Furthermore, the method includes communicating the identified and classified indistinguishable labor codes for diagnosing a root cause of real time first vehicle fault data. A computer-readable medium storing an executable computer algorithm for performing the root cause diagnosis of vehicle fault data is also envisioned.
-