The previous installment highlighted the complexity of handling unexpected incidents. This installment delves deeper into this issue from a technical perspective.
Methods for ensuring safety in automotive electronic systems have traditionally fallen under the concept of functional safety. This approach assumes that vehicles are designed to be safe. If a malfunction occurs, safety mechanisms are activated to shift the vehicle to a safe state.
For example, brake systems are designed with redundancy so that if one part fails, other systems take over and continue to provide braking functionality. A systematic design approach that includes considerations for potential failures is the essence of functional safety, and the concept has traditionally been very effective in ensuring vehicle safety.
However, in autonomous driving systems operating at Level 3 and above, the vehicle may face unforeseen situations that could lead to hazardous events, even if the system itself is functioning without fault.
It is challenging for even human drivers to anticipate and respond to all possible road scenarios. It is particularly difficult for autonomous driving systems to respond perfectly in unexpected situations.
A new concept designed to ensure vehicle safety even in the face of unforeseen factors is known as safety of the intended functionality (SOTIF). Under the framework, risk scenarios can be categorized into four areas:
While necessary measures can be taken for known scenarios, countermeasures cannot be implemented for unknown scenarios, so hazardous unknowns represent the biggest challenge. To operate Level 4 autonomous vehicles, it's crucial to minimize the number of unknown hazardous scenarios. This involves monitoring real-world events and taking necessary measures when previously unidentified risks are discovered.
Even after an autonomous vehicle begins operating in the real world, continuous monitoring is critical to identify hazardous scenarios and make improvements. Unlike conventional vehicles, this process does not end with the sale of the car. Software development involves an endless loop of refinement and updates.
Traditionally, automotive software has been developed within a closed environment based on concepts such as functional safety. However, autonomous driving development requires approaches like SOTIF, which involves ongoing adaptations that take external factors into account.
Automotive software is now at the threshold of an open environment, where applications extend beyond individual vehicle models. The next installment will showcase how TIER IV’s innovative products and solutions are pushing the boundaries of intelligent vehicle development, setting new standards for the future of autonomous driving.
Toshihide Ando | TIER IV Fellow
Toshihide joined TIER IV in 2019, where he has served in roles including vice president of technology before assuming his current position as fellow. Previous experience includes software development and R&D at a major Japanese automotive parts manufacturer.
TIER IV is always on the lookout for passionate individuals to join our journey. If you share our vision of making autonomous driving accessible to all, get in touch.
Visit our careers page to view all job openings.
If you’re uncertain about which roles align best with your experience, or if the current job openings don’t quite match your preferences, register your interest here. We’ll get in touch if a role that matches your experience becomes available, and schedule an informal interview.
Inquiries
Social Media
X (Japan/Global) | LinkedIn | Facebook | Instagram | YouTube
More