The recent Model 3 announcement by Tesla took the industry by storm and saw Tesla collecting a whopping $276 million in preorders in a matter of days. In focus in particular was the autopilot features on the new Tesla car – which meant that Autonomous Cars (a.k.a. driverless cars or self-driven cars) are finally breaching the line between concept and mainstream.
We have seen autonomous cars coming for years now. Google has been testing extensively its self-driving car technology under the “Google X” project since 2012 with overwhelmingly positive results. Google’s autonomous cars have driven well over 1 million miles (equivalent to 75 years of U.S. adult driving) with 15 accidents – with only 1 accident being the car’s fault. This one accident, in February 2016, the self-driving car attempted to avoid sandbags blocking its path and struck a bus. Ironically, shortly thereafter, Google was granted a patent on detecting buses by autonomous cars (of course, the patent had been filed back in March 2014).
Fueled by Google’s example, several other car manufacturers and technology companies have also been actively developing the technology and trying to commercialize the concept. Apple, Baidu, Toyota, Robert Bosch and Nissan have all dedicated resources and significant investments to build upon the development. General Motors’ $500 million investment into Lyft is seen by most experts as groundwork for an autonomous ride sharing network.
Though efforts have escalated significantly in the last five years, autonomous cars are not a new concept. Initial research can be traced back all the way to the 1920s when Houdina Radio Control demonstrated a radio controlled car in front of the New York public. In 1939, Futurama depicted cars which could use embedded systems in and under roadways to guide themselves. Later, during the 1960s, several universities started active research on autonomous cars technology with The Ohio State University and Stanford University being pioneers, though almost all the research at the time was focused on modifying roadways to guide autonomous cars. The 1980s saw the change of focus from modifying road systems to improving cars to be autonomous irrespective of road conditions. Mercedes-Benz demonstrated a robotic van that could run in busy city streets, making autonomous technology a distinct reality, albeit not commercially viable yet. Since then, research has progressed at considerate pace – with most car manufacturers, technology companies, more universities like Carnegie Mellon University and even DARPA, making significant advances.
The world’s largest automotive supplier by sales, Robert Bosch, collaborated with the second-largest high-definition mapping company by sales, TomTom to ensure continuous flow of high definition mapping data. TomTom’s maps are already being used in cars being tested by Bosch on highways in the U.S and Germany. Bosch will use its engineering expertise to help make TomTom’s maps more accurate and work seamlessly with data produced real time by the car using sensors. Universities have stepped up as well to collaborate with various automakers and technology companies to accelerate the research.
One of the leading research units in the field, Carnegie-Mellon University had announced a collaborative research lab with General Motors back in 2008. Carnegie-Mellon later also announced a collaboration with Uber on researching autonomous taxi Infrastructure for major cities. Toyota has also identified the importance of universities and provided a combined funding of $50 million to MIT and Stanford. In a similar vein, the Engineering and Physical Sciences Research Council (EPSRC) and Jaguar Land Rover have jointly invested up to €11 million into a project across 10 UK universities which will work at least in part to further research on autonomous cars.
Taking head of the imminent future, in May 2013, the National Highway Traffic Safety Administration (NHTSA) classified autonomous cars into five levels based on autonomous capabilities, which provides us a basic framework for pinning technological drift on a scale.
Level 0 – No Automation: Driver controls all the functions of the car. No Automation.
Level 1 – Function Specific Automation: One or more specific control systems are automated. For example, assisted braking or stability control.
Level 2 – Combined Function Automation: At least two of the control systems are automated in unison. For Example adaptive cruise control with lane centering.
Level 3 – Limited Self Driving Automation: All the critical systems like steering, brakes and throttle are automated. The car monitors conditions to require a transition back to driver control.
Level 4: Full Self Driving Automation: All systems are fully automated requiring no human intervention.
The current decade will prove transformational for autonomous cars. Toyota’s aptly named Highway Teammate and Nissan’s Level 3 cars are expected to be launched into commercial market by 2020 – but we might see Level 4 cars entering mainstream even before that – with Google leading the way. The industry on the whole is expected to hit 10 million vehicles on the road by end of 2020 – that’s a Compound annual growth rate of 134%.
Autonomous technology is also expected to expand beyond personal vehicles to public transportation. Local governments in many metropolitan cities are aiming towards automated transit systems to de-congest their cities in an effective manner. The first buses of this kind will be introduced in Switzerland in 2016. The Chinese bus company, Yutong is running similar trials and is expected to open services to public in near future.
The graph below predicts the amount of sales by level of penetration of autonomous cars into the automotive industry. Studies such as one by VTPI.org estimates that by 2070, every car being bought will have Autonomous technology installed as a primary requirement.
As with any new emerging technology, focus also shifts to patents and patent ownership trends on that technology. Research shows that technologies such as Adaptive Cruise Control and Anti-Collision Systems have the highest number of patents/patent applications filings, followed by Braking Control Mechanism and Communication Systems. Media and Sonar Systems have the least number of patent filings with only 396 and 597 patents/patent applications respectively.
The number of application filings has constantly increased from 1995 to today – with some deceleration in 2009 due to the economic recession of 2008-09 when a majority of the automobile makers saw cash flows dwindle. The graph shows a dip after 2013 as many of the applications filed haven’t been published yet but we can assume that the number of filings in 2015 will cross the 2500 mark and the positive growth in this technology would continue.
Looking at total patent ownership, as shown in the image below, Toyota, Robert Bosch and Nissan have the most patents relevant to autonomous cars with 3110, 2665, and 1169 patents and applications respectively. Volkswagen-Porsche (1140) has strong a portfolio in signaling and collision responsive systems while Daimler (961) has a strong portfolio in Collision Responsive Systems, Collision Detection and Pedestrian Safety Systems. Mitsubishi (231) has so far concentrated on signaling and Vehicle Steering Systems, while Panasonic (220) has patents on Vehicle Steering and Passenger Safety Systems.
The list of top 20 assignees is dominated by automobile manufacturers like Toyota, Nissan, and Volkswagen. Automobile suppliers like Robert Bosch, Valeo SA and Mando Corporation also have a significant number of patents in the areas of signaling and steering systems.
Alphabet (Google) which has been spearheading autonomous cars holds 238 patents/patent applications, majority of which are in V2V and V2I communications hinting that Google will rely on the same “leading from behind” approach with autonomous cars that it followed with Android OS for the smartphone industry. Tesla is also expected to play an important role in this space, but in terms of patents owned already, does not figure in this top 20 list.
Despite the small portfolio size, Tesla oft shares spotlight with Google – and has announced that the new Model 3 cars shipping in 2017 will be able to run in ‘Autopilot’ mode at least on freeways and to “Summon” itself out of its parking spot. But that’s not just why Tesla is dangerous for competition. Tesla’s product seems to be the most advanced commercially-viable implementation of autonomous car technology – seeing how the Model 3 is only expected to cost $35,000, which means the heat is really on for the other larger traditional automobile players (and also on Google and Apple) to commercialize their own autonomous car technology, and make it accessible to regular consumers, before Tesla takes the road from beneath their wheels.