15R - AMD - 2014-2015

Joint R&D project with Xerox Research Center Europe (Grenoble, FR)

AMD - Adaptive Model Diagnostics, applied to safe driving and vehicle auxiliary systems diagnostics

In the 'AMD - ADAPTIVE MODEL DIAGNOSTICS', funded by the Regione Veneto 'POR CRO PARTE FER 2007-2013 ASSE 5 LINEA DI INTERVENTO 5.1 AZIONE 5.1.1 "COOPERAZIONE INTERREGIONALE" S.A.T.E. designed and implemeted a general method for online system diagnosis. The method is based on a mathematical model of the system, based on no or qualitative a priori knowledge. The purpose of this approach is to avoid or reduce the acquisition tests required for the model parameters identification, which may be expensive or not always possible in industrial practice.

The general method was applied to two specific cases: the driving style evaluation and the vehicle cooling system diagnosis. The results obtained show that the diagnostic system is able to characterise the normal system behaviour in a sufficiently short period (typically few km) and to detect novelties in the driving style or status of the cooling system almost immediately.

The architecture of the system includes a smartphone APP, the DIAGNOSTIC APP, that processes the data collected by a Bluetooth interface connected to the vehicle OBD port. The APP then stores synthetic information both into the USER DIAGNOSTIC DATABASE and a remote server that collects anonymously the data of all users into the COLLECTIVE DIAGNOSTIC DATABASE.

By means of specific credentials and API (Application Interface Programming),a THIRD PARTY SERVER or APP can request information to the DIAGNOSTIC MANAGER SERVER about the driving style and the status of the vehicles cooling systems of all the users.