Automotive Whitepaper: Self-Driving Cars

(Excerpt)

One of the hottest subjects in the automotive industry has been the autonomous or self-driving car, with the conversation picking up speed in 2017. The challenges to getting this technology on the road and into a viable business continue to fall in several buckets: development of technology, regulatory issues, and consumer acceptance. Several automakers have promised commercialization by 2020, although nearly all are looking to initially deploy as ridesharing vehicles in limited geographies. The change from having a vehicle that offers sensors to avoid certain actions and is capable of supporting the driver to utilizing a fully driver-less car will require applying deep machine learning to the driving environment. There are an increasing number of activities focusing on developing this area. TESTING: REAL OR SIMULATED?

Real-world testing continues to grow, with new companies coming to the field. In addition to real-world testing, there is an increasing focus placed on the benefits of simulation testing. Waymo, for example, is testing about 3 million miles per day in virtual environments. For this kind of testing, these companies need video or pictures of a car driving down the road, situations to which the software can interpret and react. To this end, Tesla's latest software update for AutoPilot included asking its drivers if the company could gather the data from the cameras onboard properly equipped Model S and Model X vehicles. This method of collecting these images means that Tesla has the advantage of being capable of gathering views of numerous traffic situations and environments rapidly. Virtual testing does not replace the need to gather data from the real world, but it is a significant tool to accelerate autonomous vehicle development.

Fill out the registration form to access the full whitepaper.

(September 2017)

Access the full whitepaper

Fill out the form below and get access to the full whitepaper.

required fields