Autonomous vehicles received the largest percentage of private artificial intelligence (AI) investment for 2018 and 2019.
Global driverless car technology startups scooped up $7.7bn, accounting for 9.9% of total private AI investment.
That’s according to the AI Index 2019, a report compiled by the Stanford Human-Centred AI Institute, in collaboration with McKinsey & Company, AI21 Labs, Genpact, Google and OpenAI.
The annual report examines the biggest trends in the AI industry between January 2018 and October 2019, such as AI growth by country and the number of peer-reviewed research papers.
It found that while autonomous vehicles – also known as AVs – are largely being tested in the US, at least 25 countries are testing them.
Bumps in the road
Despite these large cash injections, autonomous vehicles are still an emerging technology with many obstacles ahead.
Carmakers such as Tesla already offer an advanced autopilot. But they remain some way off developing fully autonomous vehicles, which are also known as level 5 AVs.
Data shows that the crash rate per million miles driven in autonomous mode has increased between 2015 and 2018, although many of these incidents were minor.
Roger McCarthy, principal, McCarthy Engineering, believes that there were more autonomous vehicle crashes in 2018 than the data suggests.
“I believe more accurate coding would move additional crashes into the ‘autonomous’ category,” he writes in the report.
“Secondly, AV’s are driven, and have their crashes, under virtually ideal daytime driving conditions. When AV’s are finally tested in more adverse environments of rain, snow, and fog, I am sure the AV crash performance will degrade, as with human drivers.
“The technical challenges of keeping sensors clean and operational under such conditions remain.”
Other areas of private AI investment
Startups using AI in healthcare, such as drug and cancer studies, received the next highest private AI investment, accounting for $4.7bn, or 6.1% of the share.
Facial recognition technologies received the third most private AI investment.
Robotic process automation (RPA), in which software automates low-level tasks for workers, was the fastest-growing area for AI investment between 2015 and 2019.
The rapidly expanding industry, which includes companies such as Automation Anywhere and Softomotive, grew by more than 900% during the four-year period.