With Silicon Valley getting the constant press that it gets, it is almost certain that you’ve heard something about self-driving cars in the last few years. According to a specific report by Frog Design Consultancy, we are on the cusp of self-driving vehicles moving out of the Research and Development phase and into the open market, en-masse. At the same time, however, the authors add a caveat to this almost immediately. In this, they state that it is more probable that we will start with a more diversified market, instead of one that is dominated by autonomous vehicles. Before attempting to understand the underlying reasons for this, it is logical to assume that we need to understand the current state of self-driving vehicles, especially in the United States.
Roadblocks to Autonomous Vehicle Deployment
As of now, over 20 US states have passed laws that indicate they support the continued development of autonomous vehicles. Inside of this lies the added finding that 19 US states and 60 US cities are currently piloting self-driving cars in real-time, on well-trafficked roads. According to Frog’s team, the main goal behind such tests is to get the AI systems inside of AVs ready for commercial usage by letting them take in training data from real driving. As we know, however, this type of testing is not without its risks.
Just last year, Uber’s self-driving vehicle program was put into jeopardy due to a fatal crash related to the inattention of its in-car employee, who was supposed to have been monitoring the journey. This particular crash actually happened in Arizona which had been, up until that point, one of the most friendly states to AV programs. If this one accident is taken out of the equation, however, Arizona still has a lot going for it with regards to being a pioneer with self-driving vehicles.
The same consultancy report reported that Waymo, which is a self-driving car startup that is incubated inside of Google, has become licensed to operate in Arizona as of this January. Because of this, Frog has concluded that Arizona could be the pioneer of what are called commercial autonomous vehicles. In this particular instance, this reportedly means that customers will be able to order rides from self-driving Chrysler Pacifica minivans.
On a national level, Ford, GM, and Lyft are mentioned as other big-name firms that have released and shown bold implications for the future of autonomous vehicles. Ford’s pilot of self-driving vehicles, which has been backed by a firm called Argo.AI, is seemingly being actively used now by companies like Domino’s and Lyft. In Lyft’s case, John Zimmer, the company’s president, has said Lyft feels confident that most of its fleet will be autonomous once we hit 2021. One could surmise that this is at least partially due to their experiences with Ford’s pilot program.
In what could be an attempt to upstage Lyft as well as closer competitors, General Motors has gone so far as to predict that they will put one AV into mass production by 2019, even though the regulatory hurdles in the space are considerable at this point. Others, such as Dr. Gill Pratt, who is a professional researcher for Toyota, caution that such optimism needs to be reigned in considerably.
He even hazards to say that self-driving cars should not be widely used until it is clear that they have measurable advantages over human drivers. This, in the context of Artificial Intelligence systems, could mean that phasing out traditional vehicles is not really commercially viable or safe until enough training data is gathered to prove that AVs are the best on the road. In the very least, their performance should be equal to that of human drivers, in order to prove that AVs are truly a safe option. If this is taken to be true, then it all comes down to the Artificial Neural Networks and the algorithms that run the AVs.
When these do not function at an optimal level, which includes consistently preventing any sort of accident, then it becomes less and less of a certainty whether such a vehicle can be trusted. In future articles, we will dig further into this with the aim of prescribing a possible road to widespread adoption for AVs.
Resources:
Primary Source: http://static.tongtianta.site/paper_pdf/1166885e-9c7a-11e8-912c-00163e08bb86.pdf
Frog Design Consultancy: The Road to Autonomous Vehicle Adoption: https://designmind.frogdesign.com/2018/06/road-autonomous-vehicle-adoption/
Gill Pratt’s Opinion on AVs: https://www.businessinsider.com/toyota-gill-pratt-unveils-self-driving-plans-concept-car-at-ces-2017-1
