Automatic Model Generation to Diagnose Autonomous Systems.
Jorge Santos Simón, Clemens Mühlbacher, Gerald Steinbauer · 2015
Autonomous systems’ dependability can be improved by performing diagnosis during run-time. This can be achieved through model-based diagnosis (MBD) techniques. The required models of the system are for the most part handcrafted. This task is time consuming and error prone. To overcome this issue, we propose a framework to generate formal models out of natural language documents, such as technical requirements or FMEA, using natural language processing (NLP) tools and techniques from the knowledge representation and reasoning (KRR) domain. Therefore, we aim to enable the usage of MBD in autonomous systems with few extra burden. So doing, we expect a significant increase in the usage of MBD techniques on real-world systems.