How artificial intelligence can enable safe driving today (not in the distant future)

We are excited to announce our investment in Nauto, which is developing driver safety and driving context solutions for commercial vehicles using real time artificial intelligence based analysis of video feeds.

What problem is Nauto solving?

Long driving hours, poor public perception, and expensive insurance products result in commercial fleet managers being constantly worried about their drivers’ safety and the liability they are taking on from their fleets operating in urban and semi-urban areas.

The idea of using video to assess if the driver is following safety protocols or if he/she is actually at fault when an accident happens is not new. Dashcams — as they are called in the fleet/insurance industry — have been used since late 2000s (see picture below). The hard problem here is imbuing these cameras with real time intelligence. Local real time intelligence hasn’t been available till today because it needs powerful video processing and, only now we are starting to see chips with these capabilities from major vendors.

I found a dashcam on an Uber trip. My driver, who previously drove a taxi, said it was a norm with San Francisco taxis.

Real time intelligence is valuable because it gives the driver and the fleet manager a chance to take preventative action before an accident happens. Additionally, a major cost factor with dashcam solutions is the manual analysis of video, which Nauto eliminates using artificial intelligence (AI), thus making the solution affordable to everyone from a mom-n-pop trucking company to Comcast or UPS.

How does this fit our investment thesis in the autonomous car space?

The autonomous cars space has been receiving a lot of attention of late leading up to GM’s acquisition of Cruise automation (a Draper Venture Network portfolio company). Such interest from OEMs affirms the basic thesis in Silicon Valley around investing in companies focused on sensing and perception sub-systems of the autonomous car system. One risk factor is the slow progress towards full autonomy (level IV) which limits the value that these companies are trying to create (and hence capture).

Our thesis has been to invest in companies that aim at the autonomous car space while maintaining optionality to mitigate the risk of slow progress towards full autonomy. We like companies aiming at the big prize (the fully autonomous car) in the distant (or not-so-distant) future while building strong differentiated businesses today. In 2016.

Nauto’s AI engine matures with video from real world (not simulated) miles with simultaneous access to actions that good drivers (average professional drivers are better than average consumers) are taking, and in the long run will become the “brain” an autonomous car needs. Importantly, their go-to-market unlocks large opportunities in fleet management and auto insurance industries on the way to full autonomy of the car, fitting very well with our thesis.

We are excited for our very first investment in this space. Our team is looking forward to work with Stefan and the strong team he has assembled at Nauto.

If you want to hear more about Nauto and other investments, get in touch with us from

More reading

  1. Coverage on TechCrunch and Re/code
  2. Official Nauto press release