Techniques for considering uncertainty in use of artificial intelligence models

This invention describes a system for making autonomous vehicles safer by evaluating the reliability of their artificial intelligence (AI) predictions. When the vehicle's AI makes a decision, the system compares the current situation to the data it was originally trained on. It then generates a confidence score based on how similar the current situation is to what the AI has seen before, using specific distance measurements like Mahalanobis distance, to decide how much to trust the AI's prediction.

Why it matters: Since 2019, the demand for verifiable safety in autonomous systems has grown significantly, making robust uncertainty quantification in AI models a critical component. Advances in real-time AI processing and computational power now make such safety systems more practical to integrate directly into vehicle control.

Status: Not in forceListed as no longer active. The specific reason is not in the record we hold.
How hard to build: SpecializedHeavy software, AI training, vehicle integration

3 AI reinvention directions for this patent

See all three

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

Where this one came from

Every morning we go through millions of abandoned patents and pick out 15 worth your time. This was one of today’s.

This one

+10