In a set of recommendations for the European Commission, TNO proposes to store additional information in the event data recorders (EDRs) of self-driving cars. According to the Dutch knowledge institution, the extra info is instrumental in assessing the full chain of events in the aftermath of a crash, including observation, processing, performance and context. This, in turn, can help answer liability issues and, more importantly, help avoid accidents in the future.
EDRs, popularly known as “black boxes”, are already commonly used in self-driving vehicles. However, in most cases, they only store basic sensory data. From this data, it’s impossible to tell how the car used the observations to make a decision and whether the chosen action has been performed correctly.
As part of a European study on the safety of non-embedded software, TNO analyzed the legal and business landscape related to new advanced technologies associated with digitization and artificial intelligence. Among others, the knowledge institution provided recommendations for which information to store in EDRs for road vehicles and other AI-based systems in fields such as industrial robots and autonomous shipping. The recorded data can be used for multiple purposes, eg training, safety assessment, surveillance, causation, vehicle diagnostics, testing and development.