Vehicular Open-Data Testbed

Apple in Nov 2016 raised comments to the NHTSA regarding the need for privacy while simultaneously the need for sharing of near-collisions and collisions. This is important point as self-driving vehicles compute on massive amounts of data. However, what makes it challenging is that the computation must be performed in a distributed system composed of both human and self-driving vehicles (where the self-driving vehicles will be different manufacturers with different capabilities and behaviors).

The most straightforward approach would be for all manufacturers to share their data. Though is most likely will not happen voluntary as the companies claim competitive advantage.

At the end of the day, self-driving vehicles have shown a huge vulnerability in the ability to interact with human drivers. See Uber and Google/Waymo.

Will self-driving vehicles evolve without data sharing and can this data sharing be guaranteed to protect individual’s privacy?