Software Safety for Next-Generation Vehicles

A $2 million, four-year project is underway with General Motors Canada (GM Canada) to develop software methods and tools to help ensure the safety and reliability of autonomous and electrified vehicles. The university-industry R&D collaboration is funded by the Natural Sciences and Engineering Research Council of Canada (NSERC), Ontario Centres of Excellence and GM Canada. The academic team includes McSCert professors Mark Lawford, Tom Maibaum and Alan Wassyng, as well as University of Toronto Computer Science Professor Marsha Chechik.


With the addition of software-enabled hybrid powertrains and Advanced Driver Assistance Systems, the system design and safety processes associated with the development of software technologies for vehicles have had a corresponding increase in required effort, difficulty and cost. GM Canada’s goal is to be at the forefront of establishing model management-based techniques to address software safety and compliance with standards, and to thereby improve industry practice.


McSCert is developing software methods and tools to help ensure safety for new GM Canada products and to reduce the time and cost associated with software safety activities. This work includes modelling design and safety artefacts and their relationships with the safety case, the complete argument demonstrating the functional safety of the system. The team is also developing methods to help GM engineers determine the impact of a design change on the safety case. From there, engineers can determine which parts of the safety case can be appropriately reused.

Training of Highly Qualified Personnel

It is expected that over 20 graduate students post-doctoral fellows, research engineers and research scientists will receive advanced training in functional safety, model management and system and software design. Our HQP will review industrial system design and safety processes and will apply the latest research developments on hazard analysis, safety assurance case templates and model management to demonstrate the effectiveness of their research, learn its limitations and motivate further improvements.