Does not really cover all cases of motor insurance fraud. Also using just accel/gyro data isn't sufficient to figure out what went wrong in an accident. Appreciate the originality of thinking this as a problem domain though !

The overall problem they are trying to solve has many angles to it...further analysis and understanding of the problem needs to be evolved further.... The opportunity is huge. More research is needed to evolve the overall idea.

Business model needs to be worked more.

Most car manufacturers have on-board tracking devices these days, which provide real time GPS tracking, which takes away from the originality/innovativeness of this idea.

There is a huge market space in this sector! if is marketed and executed this would be a game changer

Need to plan more for scale and thoroughness

Product is not capturing enough information for the insurance companies. Speed only cannot be the variable captured to come to result. The product can be enhanced by capturing real time usage behavior of the driver.

Hmmmm. Not well thought out. It may know a hit but not know who is at fault. It's an incomplete solution.

It is a good problem to solve. IOT and data has been an area left unresolved in the insurance industry.