263 Self-Driving Car Startups to Watch

We mapped the transportation stack, so you don’t have to.

Taylor Stewart
Comet Labs
Published in
5 min readMay 10, 2017

--

With all the attention on self-driving cars as one of the first monumentally transformative applications of AI, startups and investors alike are jumping at the chance to be part of the action. Because this is a complex problem beyond autonomy alone — and because both incumbents and startups are still carving out their space in the rapidly changing market — there are lots of founders and engineers creating innovative technologies that should not be overlooked.

We want to give them a chance to be in the spotlight.

This map, or “stack”, is our attempt to visually transcribe our research and sourcing process for the Transportation Lab. The Lab is a 3-month startup program connecting founders building tools for the transportation industry with leading corporations, expert mentors, and key resources to accelerate customer development. In everything we do, we try to emphasize the importance of collaboration to accomplish shared goals, faster. By sharing our research with our community, we hope to learn from you about even more under-the-radar technologists who are working on solutions to transform this traditional industry.

What does it tell us?

The most immediate problem to tackle is getting self-driving cars that work seamlessly as quickly as possible. Naturally, the startups that get the most attention have been the ones that are working on creating the cameras and algorithms for autonomy.

However, autonomy is just one piece of a much larger, more complex puzzle. The fully autonomous vehicle of the future is going to require lots of new technologies across many different categories.

How did we build it?

Our goal during sourcing and due diligence was to compile a comprehensive list of all AI-powered startups — and those building enabling technology — working across the transportation value chain. We then identified the most interesting teams working on specific challenges like predictive maintenance, autonomy, and mapping. We have favored early-stage companies not often featured in the press (although there are also some familiar logos). We spoke with many of them, invested in some, and are now working with an exclusive few (Transportation Lab startups are boxed in the map).

Did we miss you? Let us know: hi@cometlabs.io

How did we select the Transportation Lab startups?

It’s a tough process. We evaluate them based on the number of other companies and/or academics working on solving the same challenge, as well as the uniqueness of the technological approach used to solve it. To the extent possible, we also focused on the team as a driving factor — particularly, the correlation between the team’s skillset and the problem they’re working to solve.

To date, we’ve accepted 10 early-stage startups into the Transportation Lab. They are tackling a breadth of problems across the transportation landscape, from mapping and autonomy to fleet management and materials characterization.

And what are they working on?

Mapping & Annotation

Both of the mapping companies in the Transportation Lab help prepare autonomous vehicles for the road, faster by allowing them to better understand their environment. Nomoko’s gigapixel cameras can capture sub-centimeter 3D models of entire cities — game-changing for autonomous vehicle simulation and training. lvl5’s crowdsourced HD maps are geared at significantly lowering the cost barrier to autonomy.

A discussion around mapping isn’t complete without talking about annotation, which is where Minds.ai comes in. Their team is skilled at both dataset development and neural network design, which is crucial to building HD maps at the scale needed to enable autonomous driving.

Autonomy

Another barrier to mass deployment of autonomous vehicles is the price of LiDar, which is one of the most expensive components of a self-driving system. AutoX’s camera-based autonomy solution costs only a fraction of the price of LiDar systems and has been successfully tested in challenging conditions such as at night and in the rain.

In addition, Oculii is developing affordable 4D tracking radar that enables autonomy and safety — like collision avoidance and blind spot detection — on a mass scale. Meanwhile, DeepVision is bringing deep learning and computer vision to vehicles and other embedded devices with their low-powered silicon processor.

On-Board Sensors

In order for autonomous vehicles to navigate city streets, in addition to highways and suburban neighborhoods, there will be little room for error when it comes to knowing where they are on the road. PointOne Navigation offers a precision GPS solution that can measure location with 10cm accuracy. Although location is their focus, they were still able to win a recent self-driving car race against some of the most established players in the autonomy space.

Fleet Management

With so much of the transportation discussion dominated by how we’ll get to full autonomy, fewer startups are focused on what happens after we get to 2020 (or 2018 or 2021 depending on who you ask). BestMile isn’t concerned with getting there, but rather how the technology will enable mobility once it’s here. Their cloud management platform for autonomous vehicle fleets sends optimized missions to vehicles, handling planning, real-time automated dispatching, intelligent routing, and electrical energy management. ActiveScaler is tackling a different piece of the fleet management challenge with their integrated, AI-powered fleet management dashboard.

Material Characterization

Finally, passenger comfort will only become more important as we transition to robotic cars. Syntouch is one of the breakout startups in this space, making a scientific instrument that can quantify human touch, helping manufacturers select the best materials to improve the in-car experience.

So now what?

Lots of startups are already working on increasingly specific parts of the self-driving car whole. While an “all-in-one” solution sounds nice in theory, we can get to better transportation solutions faster if startups and large corporations work together.

That’s why the Transportation Lab is a collaborative program involving corporate and nonprofit partners, such as TomTom, NVIDIA, and Bosch Venture Capital.

By bringing partners and startups together through the Lab, we’re creating an environment for them to test and develop solutions together to accelerate the transformation of the transportation industry.

We know this is only the beginning.

Our goal was to create a comprehensive list of startups in the space, but we know there’s startup founders and academics who we missed.

Are you one of them? Let us know what you’re working on: hi@cometlabs.io

If you thought this map was interesting, we’d love it if you’d hit the heart below to recommend it to other readers.

And we intend to iterate on this, so follow our publication for updates!

--

--