Nauto & Starsky Robotics: Anticipating Autonomous Vehicles


Despite the NHTSA (National Highway Traffic Safety Administration) and others describing a seemingly smooth glide path to autonomous vehicles (Level 1 to Level 4 and, voila, you are there) in reality there are going to be many paths, mostly rocky ones at that. Recently, we looked at two companies that are taking early steps – in very different ways – that can assist the efforts towards autonomous driving. The companies are Nauto and Starsky Robotics.

Nauto: Cameras in Cars Backed by AI

Nauto installs what it describes as an “intuitive AI-powered dual-facing camera” that can observe the driver’s conduct within a car and also collect images from the area around the outside of the car. The camera includes smart sensors that analyze driver behavior, monitor vehicular activity and track environmental conditions.

In a recent discussion with the company, they emphasized that they are addressing markets, initially of fleets of professional drivers. They are also licensing their technology to auto companies who will hopefully integrate it into some of their models. Two of these OEMs were identified in October 2016, when the company announced an investment by Toyota, BMW and insurer Allianz, which also included licensing the Nauto technology and data.

Nauto, Expanding AI

The images collected by the cameras obviously need interpretation, which is the role of the AI. The information can be useful in a number of contexts. On an immediate basis, the system is due to be upgraded to the point where it could warn the driver of possible risks, such as a possible collision, or coach them, for example on a route to take.

On a longer term basis, the information is fed into Nauto’s cloud and a Vision Enhanced Risk Assessment (VERA) of the vehicle is produced for the fleet manager, or an insurance company, or the driver. Over the very long term the accumulation of data can be used to derive statistics about driver risk, traffic management, accident causes and innumerable other aspects of driving and traffic.

Nauto: The Bigger Picture

CEO Stefan Heck makes a strong case regarding the waste and inefficiency in the ground transportation marketplace. We have commented several times on the fact that the automobile is probably the worst investment the average consumer ever makes, because it is used only about 5% of its lifetime and a high percentage of that “usage,” as Heck points out, turns out to be looking for parking spaces.

Heck’s message appears to be not so much a commercial one, as one of appealing to society – meaning in most cases, governments – to make improvements by mandating management of peoples’ driving habits (or eliminating the people altogether), in the name of economic efficiency.

Obviously there are numerous alternatives available for in-car cameras. So the key to Nauto’s strategy has to be a race to build up its penetration of vehicles and its accompanying databases, built around its AI core of intelligence.

Starsky Robotics: Remote Controlled Trucks

Starsky Robotics claims: “We make robots that automate vehicles.” The concept is that a truck drives autonomously on the highway but is subject to remote control by a driver for the first and last mile (getting to and from the highway.)

The company installs its robotic controls in the truck cab. These controls can turn a steering wheel, push gas pedals, brakes and clutch, manipulate gear shifts. The robotics can be controlled by a driver sitting at a remote computer with a steering wheel (somewhat similar to a computer game.)

The company’s mantra is that they are keeping the drivers employed, but getting them out of the cab of the truck. They believe that their drivers can monitor several on-road vehicles at the same time. The drivers will take over when the truck is entering or leaving a highway.

Starsky’s Differentiator

The truck is thus also equipped with sensors and software for self-driving. Obviously, there are other companies, such as Otto (owned by Uber) and Embark, supplying self-driving truck capabilities. Starsky’s differentiator is the use of remote controls.

Starsky’s approach has interesting implications. First, it means that the truck is always under human surveillance as it proceeds on its journey. Secondly, it means that a number of tasks, beyond long-haul driving, can also be accomplished remotely. An excellent example is moving truck cabs around a trucking yard, which would ordinarily take the time and energy of drivers. This can be done more efficiently by remote control.

Our Take

Both of these companies are interesting because they are offering solutions that are retrofitted to the vehicles they are serving. In other words, in contrast to many players who are edging their way along the long paths to semi- or fully- autonomous vehicles, Nauto and Starsky are attempting to build businesses that service today’s vehicles.

In the wild acquisition atmosphere surrounding anything that has the slightest relationship to autonomous vehicles, they both appear to be promising acquisition candidates, assuming their funding holds out long enough to allow them to gain market traction.


Regarding Nauto, we looked at a similar type of application in a quite different industry, namely Fast Food. (“EZUniverse – Brilliant Mobile Cloud Service/App in Fast Food” MCE 8/7/16.) It’s obvious that video is a tried and true, logical medium for capturing and analyzing events, whether it’s an employee dipping into the till at a Burger King, or a cab driver ogling girls while driving the streets in Beverly Hills. The value add, however, is in the ability to interpret the visual data quickly and accurately and transmit it efficiently to business decision makers.

The role of video in the long term development of autonomous cars is uncertain. A recent article on SeekingAlpha discussed the use of LIDAR (Light Detection and Ranging) by Google’s Waymo and others, versus Tesla’s use of an array of cameras (and sensors and radar). It speculated that Tesla’s may be the better approach for the long term. (“Alphabet Offers A Glimpse Of The Self-Driving Future” SeekingAlpha 4/25/17.) This is but one type of issue to be sorted out on the long, winding path to autonomous vehicles.

Starsky Robotics

Starsky’s approach implies a number of questions about the future, not only of autonomous trucks, but of all vehicles. The idea of some remote monitoring and control is one that we believe is quite intriguing. As we go down the long path to self-driving vehicles, it will be interesting to see how many types of situations remote monitoring and control may be applied in and its impact on factors such as safety and security.

Visit their websites: