"Toyota's [entire fleet of vehicles] drives 1 trillion miles per year or about 10,000 miles per car," Pratt said. "How many accidents caused by some error of the machine, let's say software, will the public put up with? The answer is almost none."
The standard that the TRI is trying to beat with its guardian angel ADAS technology is that of humans' natural ability to avoid crashes, which is a high standard. Most important, he said, is to avoid adding any technology that might cause a crash instead of preventing one.
Some of the ADAS technology will be passive, as in today's warning systems that beep or vibrate the steering wheel to alert drivers of hazards, while others will be active, such as self-braking cars.
The U.S. Department of Transportation's National Highway Traffic Safety Administration (NHTSA) and the Insurance Institute for Highway Safety (IIHS) announced earlier this year that 20 automakers have pledged to make automatic emergency braking (AEB) standard on their cars by 2022.
A lot of the discussion among automakers and within their R&D organizations involved how much control the car should have.
For example, Pratt said, your car may someday warn you several times about a particularly dangerous driving habit you have before taking control of the wheel.
Autonomous driving capabilities are measured on a government scale of zero to four, with zero being no automation, and four being fully automated.
The focus of most of the discussion among car makers today is how far up the scale they should go and how quickly, Pratt said.
"There's a lot of discussion in the industry whether we go incrementally up the scale or whether we jump," he said. "The chauffeur and guardian angel are built out of the same technologies. I actually believe they are complementary to each other."
Sign up for CIO Asia eNewsletters.