Vehicles equipped with sensors and cameras navigate the streets of Mountain View, California; Austin, Texas; Kirkland, Washington; Dearborn, Michigan; Pittsburgh, Pennsylvania; Beijing, China; Wuhu, China; Gothenburg, Sweden; Rotterdam, Netherlands; Suzu, Japan; Fujisawa, Japan; and Seoul, South Korea, among other places. Sophisticated on-board software integrates data from dozens of sources, analyzes this information in real-time, and automatically guides the car using high definition maps around possible dangers. 

People are used to thinking about vehicles from a transportation standpoint, but increasingly they have become large mobile devices with tremendous processing power.2 Experts estimate that “more than 100,000 data points” are generated by technology in a contemporary automobile.3 Advances in artificial intelligence (software that applies advanced computing to problem-solving) and deep learning (software analytics that learn from past experience) allow on-board computers connected to cloud processing platforms to integrate data instantly and proceed to desired destinations. With the emergence of 5G networks and the Internet of Things, these trends will harbor a new era of vehicle development.

Between now and 2021, driverless cars will move into the marketplace and usher in a novel period.4 The World Economic Forum estimates that the digital transformation of the automotive industry will generate $67 billion in value for that sector and $3.1 trillion in societal benefits.5 That includes improvements from autonomous vehicles, connected travelers, and the transportation enterprise ecosystem as a whole.

Eoin Treacy's view

Here is a link to the full report.

There are two very big questions when it comes to the viability of self- driving cars. The first is whether it is technologically feasible to let a fleet of autonomous vehicles loose on the roads where the actions of unpredictable pedestrians, animals and weather will test an artificial intelligence to the limit. The second is the extent to which governments will successfully regulate for these vehicles so that insurance considerations can be ameliorated. 

The first is in many respects the easy part. Trials are underway in Singapore and Pittsburg to trial new autonomous vehicles run by companies like Uber and Lyft. Speculation today that Apple has opened a conversation with McLaren on possibly taking a stake suggest the company has its eye on the automobile market. Artificial intelligence engines only need to learn a lesson once and thrive on the quantity of data being picked up by all of the test vehicles circulating metropolitan areas. If they have problems today none are so large that they cannot be fixed by additional iterations of the software. 

The second is a potentially more thorny subject because until now drivers are personally liable for their actions behind the wheel and pay insurance companies to mitigate that risk. The rules imposed by the USA insisting that autonomous vehicles must have a steering wheel, brakes and a licenced driver at the helm are all designed to ensure someone can be found at fault in the event of a crash. It is going to require some inventive actuarial work to come up with a formula that will allow companies develop passenger vehicles that will remove personal liability from drivers. That suggests the evolution of the autonomous vehicles is likely to take longer than some of its biggest cheerleaders project.