A vehicle is designed to decide whether to perform a specific action, such as driving along or maintaining a particular path, only if it's sufficiently confident in the accuracy of the information its internal systems are providing for that decision. If the vehicle's systems are confident enough about a proposed action, like staying in its current lane, it will proceed with that action.
Why it matters: Filed before the widespread maturity of AI/ML models in autonomous driving. The ability to generate and interpret robust confidence scores from these models for critical decisions like trajectory planning has significantly advanced, making this approach more practical for enhancing vehicle safety and reliability.
AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.
Reinvent this with AI