Find us on...
- Products & Services
- Sectors & Roles
- Report Store
- Contact Us
- Request a demo
Billions of test miles to prove driverless tech
As of April 2016, Google is running 23 converted Lexus SUVs and 34 prototype vehicles in autonomous mode on public roads across Kirkland, Mountain View, Phoenix and Austin in the US. It has driven its vehicles a total of 1,564,981 miles in autonomous mode to date, averaging weekly miles of between 10 and 15 thousand.
On May 3rd 2016, it was announced that Fiat Chrysler Automobiles and Google will be joining forces to integrate autonomous driving technology into 100 Chrysler Pacifica Minivans, marking the first time that Google has partnered directly with an automotive manufacturer. The new vans will be used to increase the amount of self-driving testing that the company can carry out in order to improve the systems it uses and help to allay any safety concerns the public may have with autonomous technology. Engineers from both teams will therefore be working together closely to accelerate the adoption of driverless cars.
The three key potential stumbling blocks to public adoption of autonomous vehicles are: technological barriers, regulatory issues, and consumer confidence. The technological barriers have been worked on for many years but it is only recently that regulatory and consumer concerns have come to the fore. Google has been lobbying hard to ensure the opening up of the legislative environment and has successfully gained testing licenses in a number of US States. Elsewhere, the UK government is set to allow driverless car trials on UK motorways from 2017. Such initiatives clearly open the door for developing legislation for full public use of autonomous vehicles.
Satisfying public perception and consumer confidence may end up being the hardest task to complete. A recent analysis conducted by RAND Corporation finds that the level of proof required to demonstrate to a reasonable degree of confidence that an autonomous vehicle is safer than the average human driver in terms of on the road fatalities is so high that it could take over 500 years of testing before full satisfaction is obtained.
Whilst the RAND model can be updated to include the addition of the 100 Chrysler Pacifica test vehicles and adjusted to be in line with real world average driving speeds, the challenge of proving the superior safety of autonomous vehicles remains almost insurmountable. Despite the marketing claims of various manufacturers that autonomous vehicles will be safer than human driven cars, it is therefore likely that autonomous vehicles will be made available for public use prior to that claim being statistically validated.