By: Milad Mohammadai
Driverless cars have survived the road, but will they survive Congress?
With the recent crash involving a Uber operated self-driving car in Arizona, driverless car technology has again come into question. Having been all the craze over the past few years, this technology has had its staunch proponents and fierce critics. Tesla’s “Autopilot” project has been a strong example of this developing technology being rolled out on a more widespread level. However, as Tesla continues to develop its driver assist technology, new entrants in the market including Waymo (Google subsidiary) and even an old giant, Ford are right on Tesla’s heels. Recently however a new project has emerged, managed by a former legend of the hacking world, George Hotz. His project Comma.ai, is an entirely new take on the self-driving car. Instead of approaching the technology from a integrated software and hardware approach as Waymo and Tesla have done, Comma.ai focuses primarily on developing an open source software based system for currently operating ordinary vehicles. Unlike Tesla’s projects, Comma.ai is designed to adapt to existing vehicles in production, with the end goal of having a machine learning based software suite that would permit users to inexpensively use their current vehicle as an autonomous vehicle. Comma.ai’s method of reaching this goal is to allow anyone to have access to their source code, and permitting open source development. The software takes advantage of sensors, modules, and other already existing hardware components in vehicles to enhance safety and usability. The software is designed to be run on most smartphones, along with a special attachment.
This new approach will likely have a drastic effect on the market, as it would permit the development of self-driving car technology by a much wider developer base as well as those with less financial capital. If this open developer platform takes off, it is likely to further accelerate adoption of self-driving car technology through its impact on the passenger car market, hired transport market, and the logistics industry. The only question now is how will the market and Federal government react? There are likely three major markets that will be affected. The first would be the traditional passenger car market, specifically private drivers. The second would be hired transport such as taxis and buses. Third, would be the logistics and trucking industry. There are roughly 1.7 million drivers in the logistics industry and another 1.7 million in the taxi and bus industry. While numbers on the revenue generated by the taxi and bus industry in the US are murky, the American Trucking Association has found that for the trucking industry, the total yearly revenue was 676 billion dollars in 2016. In the taxi industry, which employs over 143,000 licensed drivers in New York City alone according to a 2016 report by the Taxi and Limousine Commission, pressure is already very high as a result of the introduction of ride-sharing services such as Uber and Lyft. Likewise widespread adoption of driverless technology, while still a distant future, will have drastic effects on this industry’s employment. Because of this, new proposals in Congress are already focused on creating new regulatory regimes for this technology. Proposals include preempting state regulations (of which vary from state to state), allowing for vehicles to be tested that don’t meet Federal safety standards, and testing caps. However, one of the advantages of the Comma.ai project is its ability to be adapted by many more users, which would likely mean regulators could do very little to prevent the spread of driverless cars. Their only choice will be to embrace it. Since the Comma.ai project is software based and is not tied to any particular car manufacturer, users are free to test the software on potentially any compatible vehicle. Because of this, it is preferable for the Federal government to take a more hands-off approach. By attempting to regulate and control the inevitable, Congress will inadvertently make it more difficult to safely adapt this technology for widespread use.
While there are many valid concerns over safety, studies and reports have demonstrated that even at their early alpha stages, self-driving vehicles are largely as safe, and if not more safe than humans. Some studies such as one conducted by the RAND Corporation even recommend adopting such technology as fast as possible, since it has the potential to drastically reduce traffic fatalities. In a predictive model where a 10% safer than human vehicle is compared to a 75% safer than human vehicle averaging development wait times, RAND Corporation’s Kalra et al. finds that “In the short term, more lives are cumulatively saved under a more permissive policy than stricter policies requiring greater safety advancements in nearly all conditions, and those savings can be significant — hundreds of thousands of lives”.
Regulating an early market by instituting testing caps may only lead to the prevalence of less than desirable systems in our current market, and may even prevent a potential safety enhancing network effect if more self-driving vehicles are permitted on the road. It remains to be seen what the future holds, and more importantly how this fast evolving technology will affect the valuation of both the passenger and commercial transport markets.
American Trucking Association. (2017). Trucking Industry Revenues Were $676.2 Billion in 2016. Arlington: American Trucking Association. Retrieved December 2017, from http://www.trucking.org/article/Trucking-Industry-Revenues-Were-$676.2-Billion-in-2016
Coma.ai. (2017). Comma.ai. Retrieved December 2017, from Comma.ai: https://comma.ai/
Conger, K. (2017, October 9). Self-Driving Cars Are Super Duper Safe, Self-Driving Car Companies Say. Gizmodo. Retrieved December 2017, from https://gizmodo.com/self-driving-cars-are-super-duper-safe-self-driving-ca-1819292527
Marshall, A. (2017, July 19). Congress Finally Gets Serious about Regulating self-driving cars. Wired. Retrieved December 2017, from https://www.wired.com/story/congress-autonomous-self-driving-car-regulations/
Joshi, M (2016). 2016 TLC Factbook. Retrieved from. http://www.nyc.gov/html/tlc/downloads/pdf/2016_tlc_factbook.pdf
Nidhi Kalra, D. G. (2017). Estimating the Cost of Waiting for Nearly Perfect Automated Vehicles. Washington D.C: RAND Corporation. Retrieved December 2017, from https://www.rand.org/pubs/research_reports/RR2150.html
Rocha, K. (2017). Road Tarmac Mountains Car. Pexels. Pexels. Retrieved December 2017, from https://www.pexels.com/photo/road-tarmac-mountains-car-57645/
Waymo. (2017). Waymo. Retrieved December 2017, from Waymo: https://waymo.com/