FaultFinder

6 million roadway incidents occur every year. Each incident induces stress and takes away hard-worked dollars. FaultFinder leverages common sense reasoning and suggests fault within a car accident.

Comment

Inspiration

Motor vehicle crashes are one of the highest causes of death in Americas and NHTSA report that roadway incidents cost Americans $340 billion in one year. Roadway incidents are inconvenient and dangerous for every party member involved. The investigative process can be long and tedious when it matters, such as when a victim in a car accident requires their medical expenses paid for after a traumatizing accident. FaultFinder leverages automated reasoning to help suggest who is at fault when an accident occurs and an unbiased judgement of who is at fault is intelligently created.

What it does

Given facts about 2 vehicles it determines who is at fault.

How we built it

Using our web interface we can generate facts for s(CASP). Then, we can use a s(CASP) program to determine whether vehicle 1 or vehicle 2 is at fault using common sense reasoning. We are able to stimulate an intelligent unbiased human being.

Challenges we ran into

The biggest challenge we ran into was the edge cases(different for each accident) and also to weigh each fact. Here are 2 examples to illustrate:

  • Car accidents run into interesting edge cases, for example, when there is a rear end accident. the vehicle that hits the car in the front is at fault unless the car in front sudden brakes(edge case).
  • Let's say the driver in vehicle 1 was drunk driving, and out of lane and the driver in vehicle 2 was distracted driving, Speeding >25mph and no turn signal. How would we weigh these facts using logic programming and determine who is at fault.

Accomplishments that we're proud of

We're proud of the fact that we got as far as we did in as little time as we did. From an idea to an MVP, we created a product with high potential from scratch.

What we learned

We learned how to use s(CASP) to implement automated reasoning models.

What's next for FaultFinder

What's next is to increase the constraints, add more edge cases, and nuance so that it can be used for more scenarios, and to ensure fault is correctly placed. We also need a good priority to determine which facts are more at fault than others. I believe that this project can be taken further and maybe be used in commercial insurance to determine who is at fault in car accidents.

Built With

Share this project:

Updates