Tesla and Google are both pushing hard into the autonomous "self-driving" car space, but the two companies are going down very different roads.
By now everyone has heard about the crash of a Tesla vehicle in May that killed a driver who was using Tesla's autopilot feature. Self-driving cars are going to crash periodically of course, and while sad, that fact is unavoidable. There are hundreds of thousands of people killed annually in car wrecks, and if self-driving cars can reduce that figure at all while simultaneously making driving easier for millions, then it's a great technology. The issue at hand is what is the best route forward for that tech?
The road thus far has taken Tesla and Google down two increasingly divergent paths. Tesla's approach has been to make its autopilot feature as limited as possible while still being a technology improvement, and then roll it out to the public for beta testing. Google's approach is to make its autopilot as advanced and different from current vehicle operation as possible, then keep the technology in house. To date, the public cannot use Google self-driving tech, nor is there any plan in the foreseeable future to make the technology available to the general public. As a result, at this point there is no way to know whose technology is safer - Google's or Tesla's.
Self-driving technology is rated by application on a one to five scale where a one is essentially just simple parking assist tech and a five means that a human functionally can't do anything to the car while it's operating other than just sit in it. Tesla's autopilot feature is rated a two. It will let the driver turn over control of the car on the highway and a computer will take over lane changes, keeping up with traffic, etc. Google's technology is a four or a five; at most humans can put in an address and select navigation directions with a GPS. The car does the rest. Related: Saudi Aramco IPO Has Investment Banks Flocking to Riyadh
It is easy to critique Tesla's tech as not being advanced enough, but the reality is that the company is working with the state-of-the-art computer systems. While Tesla's system did lead to a crash and it can be faulted for that, it's impossible to know how many other crashes the system may have helped avoid in the past.
Moreover, Google's system likely would have had exactly the same problem. The Tesla crash was not caused by a human interfering with the car, but rather by a failure in the software. Google's technology would have had the same issue. The alternative is to let humans make decisions rather than turn the driving over to a computer - but of course that method is the system 99.99% of the world relies on today and crashes happen all the time due to human error.
Broadly speaking there is no doubt that Tesla is taking the less risky route by trying to make a small but important change in the way people drive today. That fits with a technology that is actually being used on the road. Google's technology makes more sense for a company that hopes to make all existing cars obsolete and upend the way the world travels. That is appropriate for a company that is keeping its approach in house for years to come.
At the end of the day, both Tesla and Google are being logical about their approach to self-driving cars. Only time will tell which system works better and which company comes out on top, but for now it's worth cheering both company's systems along and hoping for continued improvements in the future that make both systems safer and safer.
By Michael McDonald of Oilprice.com
More Top Reads From Oilprice.com:
Michael is an assistant professor of finance and a frequent consultant to companies regarding capital structure decisions and investments. He holds a PhD in finance… More
Comments
Google's self driving car is using LIDAR radar detection ... there no such thing as being blinded by the sun. This crash would not have happened if Telsa was using LIDAR instead of ordinary cameras.
In the meantime, I like using AP. I accept that it is my responsibility as the driver to remain in full control of the vehicle at every moment, but it does provide a layer of responsive safety to what I can maintain. And it reduces the tidium and stress of manual driving. Tesla finds that the rate of accidents per mile (measured by airbag deployments) under AP is about half that without AP. So a driver like me is contributing both AP and manual miles with the same car, same driver, same road, and many other comparable conditions. This is a pretty good basis for fleet learning. So the early indication does suggest that AP is reducing the accident rate for Tesla drivers. It will take a few more years to accumulate the 1 billion miles needed for statistically significant results. But Tesla drivers are logging this data every day, and I am personally proud to contribute to this.
Expect sales of new combustion engine private cars to be rare within 20 years. The increasing pollution taxes, and very high cost of gasoline by then, will make them toys of the rich, if they are even still legal in developed countries.