Tesla: No More Driving
Where is my steering wheel?? Tesla changing the way we drive or not drive…
The self-driving car revolution is here and Tesla is jumping feet first with its autopilot feature. Amongst its ranks include the likes of familiar brands such as Google, Uber, BMW and even Apple—but the competition is fierce and the pitfalls are many. Already Apple has started to think twice about its self-driving initiative (code named Titan) and has been forced to make significant headcount and budget reductions[1]. Meanwhile at Tesla, a much more somber episode has been triggered by the death of Joshua Brown who died while the car was on autopilot and was unable to distinguish between the sky and white body of an 18-wheeler. There is significant doubt, however, that the driver was fully paying attention to the task at hand and was instead viewing a Harry Potter movie[2]; this is especially important as the current “autopilot mode” Tesla offers is not truly a full self-driving feature.
But the fact of the matter is that this raises two important questions about the intersection of technology and human capabilities: Are we ready to trust our cars to drive us? And are we ready for the change that will invariably come when human drivers are no longer needed? As a matter of trust, self-driving car technology to date has proven remarkably safe [3] in limited contexts especially compared to human driving. With tens of thousands of motorist deaths every year in the United States alone, there is a great opportunity for saving lives (not to mention eliminating drunk driving). Or as Elon Musk, illustrious CEO of Tesla puts it “[You’re] killing people,” by halting the progress of this technology[4].
The technology of course is extremely sophisticated. Many of the self-driving software and hardware today rely on being able to build a 3D map of the car’s environment and then making decisions about how to navigate that environment. This often involves a combination of high powered sensors including GPS, LIDAR (Light Detection and Ranging), and RADAR, coupled with complex algorithms and programming software that computes and makes driving decisions. These all come with caveats and have only been tested on a limited number of road types and conditions to date. For example, whereas a human can operate a vehicle while snowing or stop when signaled by a police officer, self-driving cars cannot (yet).
Despite these challenges, there is an enormous potential for autonomous vehicles to change life as we know it. Long commutes will be a thing of the past since theoretically you could eat breakfast and have a nap while your car drives you. Need to get away? Go to bed in your car and wake up 8 hours later. Think about all the transportation inefficiencies, parking nightmares, and cluttered roadways—will these be a thing of the past? But all this disruption might spell disaster for the millions of workers who currently make their livelihood driving things around. If driverless transport is the new norm, you can guarantee that truck drivers, taxis, movers, and the entire shipping industry will be impacted. Imagine a pool of intelligent cars—which may or not may not look like what they do today, since they have no need for a “front” or a steering wheel—whizzing around able to be summoned for whatever reason. Perhaps this is the new future, a future where we will wonder how we even let humans operate large heavy machines at high speeds with barely any distance in between! Of course the clear winners in this game will be those companies able to provide the best operating software for the cars [5]. With so much disruption there is a lot of money to be made after all.
As far as Tesla is concerned, Elon Musk has already made the bold claim of being able to get a car that can drive itself across the country by the end of 2017[6], making Tesla the first mover in this frenzy. Meanwhile regulators including the Department of Transportation have already jumped on board issuing their own policies and guidelines[7]. The stage is set as cars Tesla is making right now are coming equipped with additional hardware to allow for significant self-driving features. While predictions place mass autonomous vehicles well in to the future, Tesla is making the driverless world a reality today. And frankly, if we don’t get on board, these cars might leave us behind.
Word Count [726]
References:
1. Wakabayashi, Daisuke, and Brian Chen. “Apple Is Said to Be Rethinking Strategy on Self-Driving Cars.” New York Times 9 Sept. 2016: B1. Print.
2. Levin, Sam, and Nicky Woolf. “Tesla Driver Killed While Using Autopilot Was Watching Harry Potter, Witness Says.” The Guardian. Guardian News and Media, 01 July 2016. Web. 18 Nov. 2016.
3. Montenegro, Robert. “Google’s Self-Driving Cars Are Ridiculously Safe.” Big Think. N.p., 07 June 2015. Web. 18 Nov. 2016.
4. McGoogan, Cara. “‘You’re Killing People’: Elon Musk Attacks Critics of Self-driving Cars.” The Telegraph. Telegraph Media Group, 20 Oct. 2016. Web. 18 Nov. 2016.
5. “Who’s Self-Driving Your Car? Autonomous Vehicles.” Economist (US) 24 Sept. 2016: n. pag. Web.
6. Stewart, Jack. “Tesla’s Self-Driving Car Plan Seems Insane, But It Just Might Work.” Wired.com. Conde Nast Digital, 24 Oct. 2016. Web. 18 Nov. 2016.
7. NPR Staff. “Regulating Self-Driving Cars For Safety Even Before They’re Built.” NPR. NPR, 20 Sept. 2016. Web. 18 Nov. 2016.
8. Inside Edition. “See Motorists Play, Read, and Relax in Self-Driving Cars as Second Tesla Crashes.” www.youtube.com/watch?v=qnZHRupjl5E, YouTube, 6 July. 2016. Web. 18 Nov. 2016.
One of the biggest difficulties I foresee is pushing past our bias against machines. Brown’s death is a case in point. Tens of thousands of Americans die every year due to human error in driving and this has become a banal norm. But when a machine error kills one person, it becomes national news. I suspect we’re already at the point that having Google’s technology do most of our driving would make the roads safer, given just how bad people are at driving. What’s needed as self-driving cars improve is probably an awareness campaign to demonstrate just how well a Model S compares to a human in consistency, focus, alertness, reaction time, maneuver precision and all the other crucial aspects of driving. I have faith in the machines- in people, less so.
Many of the positives of self-driving functionalities of personal automobiles mirror that of public transportation (reading/sleeping during your trip, parking, reducing congestion issues). Perhaps, at the moment, the main advantage is that Tesla and other personal self-driving cars are not bound by a schedule whereas individual’s travel times are dictated when there are buses, trains, etc. However, even Tesla’s long term plan includes options for mass transportation (aka bus). I would be very interested in seeing how self-driving features and public transportation may converge.
An interesting component of the driver-less car debate is the moral issues associated with these cars. While I completely agree that in aggregate self-driving cars will save lives, there will still be situations that machines will make choices that determine whether passengers or pedestrians are more likely to be hurt in a dangerous situation. An article in Science entitled “The Social Dilemma of Autonomous Vehicles” from earlier this year outlines this debate and comes to a similar conclusion that your post did in saying “regulating for utilitarian algorithms may paradoxically increase casualties by postponing the adoption of a safer technology.” (http://science.sciencemag.org/content/352/6293/1573)
I am curious to see how the public perception of self driving cars evolves as their popularity grows. Right now, with few driverless cars on the road, the public seems to be enthusiastic about the idea of not having to get behind the wheel of a car. But as more and more driverless cars accumulate and inevitable accidents grow, the perception might dwindle. Hopefully driverless cars do become a widely accepted thing as I would argue that the greatest benefit to driverless cars is reduced traffic.
https://www.youtube.com/watch?v=Suugn-p5C1M
The above video shows how a group of cars were instructed to drive along a circular track at a constant speed of 30 km/hr. After some time, the cars begin to get closer together and traffic jams build up. As we’ve seen in earlier TOM classes, this is due to the variability of the driving rate of each car. In 100 years from now, when hopefully all cars are driverless, this variability will be close to zero (assuming cars follow the normal rules of traffic). In that case, traffic not caused by an accident of physical delay will be minimized.
I completely agree with David’s comment above. Assuming driverless technology continues to progress, the next big hurdle will be bias and regulation against autonomous machinery. One comparison to this change is elevators. Elevators used to be operated by humans and for good reason. They were not perfect machines and they made mistakes that cost people their lives. However, when elevators became more automated and, statistically, safer than their human-operated counterparts, people still did not trust them [1]. In addition, elevator operators went on strike. It took time and convincing to get people comfortable with the idea of automated machines. I predict the same barrier with driverless vehicles.
[1] http://www.npr.org/2015/07/31/427990392/remembering-when-driverless-elevators-drew-skepticism
Nice post Atul!
I’ve spent a decent amount of time thinking about the future of driverless cars so this article was right up my alley. Particularly, your post brought to mind this in-between space we’re in today where a) Teslas can drive themselves, but need to be ‘monitored’ and b) self-driving cars share the road with human-operated cars. Predictions of when autonomous cars will rule the streets vary dramatically and I think once we can zero-in on consistent rules for transit, we’ll have a much better idea of when this future becomes possible. For instance, I think that driving on the freeway could be done completely on ‘auto-pilot’ but once off these designated freeways, humans take control and do so fully. As human beings we pay attention poorly, and need to know exactly when we must be focused vs. when we can space out.
Almost every country song has a line about driving your big red truck on the back roads and tasting that pure, American freedom. I wonder to what extent people outside of our bubble actually think self-driving cars are an infringement on their freedoms, and thus limiting adoption on the consumer retail side. Giving up control to a machine, especially machines that are replacing good paying, working class jobs, doesn’t seem likely in the short term. I can see there being significant push back from consumers who see their cars as a piece of their identity, an experience rather than a tool. With self-driving cars, we’re assuming that people want (or need) extra time in their day and that driving is an inconvenience. Certainly, people don’t get in their cars thinking “I’m going to crash this baby today,” so safety isn’t a problem we’re solving at the individual driver level. I think these assumptions aren’t universally true and will lead to an uphill battle for technology adoption, unless the government specifically mandates that cars must be self-driving. I could even see this being as contentious an issue as gun regulation – it’s similarly irrational and deeply routed in the idea of American freedom… we’ll see!
Great post Atul! It has been fun to watch the strides Telsa has made in self-driving cars. Do you think this is a winner-take-all market (i.e. once Telsa does this, they will license or lease the technology out to other OEMs) or something that is going to become commonplace in the future? It sounds like a lot of companies are racing to perfect this technology – you mentioned Google, Apple, Uber, and BMW among others. While the benefit to a company like Uber or Google is quite obvious – will having an “autonomous” feature for an OEM really move the needle in terms of car sales? I feel like most people want the assurance of being able to drive their own cars and may be hesitant to adopt such technology until it is widespread and endorsed by the government. Speaking of – what role will government regulation play in the future of autonomous cars? Once again, great post and cool topic!
Very interesting post! I definitely agree that self-driving car is a technology which can completely change people’s lifestyle. However I feel that it is not easy for that technology to become a norm and accepted by everyone; several remaining problems such as technical issues, people’s negative perception against machines, protests from industries such as taxi or shipping, etc. have to be solve. How do you think Tesla will overcome all the issues and take the initiative in the car industry revolution?