Uber’s Bet on Self-Driving Car
Uber, the tech firm that invented the popular ride-sharing service and disrupted the way the world moves, is now facing its biggest challenge since its inception: the revolution around self-driving car.
Ever since the first car was invented 130 years ago, innovations in the automotive made the vehicles cheaper,safer and more efficient. However, these changes were incremental. People still drive the vehicles in the same way as in the 19th century, which has huge social and economical consequences. According to a 2015 report from NHTSA, 94% of crashes were due to human error and led to a loss of 30,000 lives and $190 billion in U.S. each year  .
As the technology evolves rapidly in the early decades of the 21st century, many companies started developing self-driving car as a solution to many issues caused by human drivers.
Company Target Timeline
Google Fully autonomous vehicle 2020
Baidu Fully autonomous vehicle 2018
Tesla Assisted and autonomous vehicle 2015
BMW Autonomous vehicle 2021
Ford Autonomous ride-hailing service 2021
Volvo Autonomous vehicle Later 2017
Beyond the improvement in safety, self-driving car could redefine the world in many ways:
- Car insurance may decrease as the accidents drop and the industry’s focus shift from millions of consumers to the manufactures of self-driving car.
- Self-driving car can free up 50 minutes a day for users  and promote in-car entertainment or enhance productivity.
- Road congestion and vehicle pollution will be reduced as the self-driving car is more efficient.
- Freight and delivery industry can benefit directly from reduced labour cost and create new opportunities for lean supply chain and low-cost delivery service.
- The development of self-driving car will promote the adoption of other robotic consumer and industry applications as they share many core technologies.
However the rise of self-driving car can completely disrupt the ride-sharing service Uber currently offers.
Current Business Model
Uber works as a two-sided platform connecting drivers with riders. It does not own the vehicles and relies on individual driver driving their own cars. The drivers are working as contractors so Uber has less expense. Because Uber is a two-sided platform and is light on assets, it built a strong network effect and enabled rapid expansion into new markets. Unlike car rental business where different companies serve different market and geography, the ride-share service tends to be highly concentrated and leads to a winner-take-all situation.
One of the biggest impacts of self-driving car, however, could be the elimination of private car ownership. On average a private car has only 5% utilization and is parked in the garage or parking lot when not driven. Around 20% of commercial land in U.S. metropolitans is used for parking, a huge waste of resources. When self-driving car is applied to ride-sharing service, it could reduce the ride fare to 1/3 or even 1/5 of current level if we factor in the driver salary and improved utilization. This would make private car ownership extremely expensive in comparison. As more people stop own private cars, Uber’s business model as a two-sided platform break.
Future Business Model
Therefore Uber needs to build and own their fleet of vehicles in order to keep the cost low and maintain a large vehicle network. Uber can partner with existing car manufacturers to integrate sensors and computer system. Since Uber controls the huge rider user base, it can demand favorite deals with car manufacturers on the ownership of user data and revenue sharing of in-car service. Uber could also develop the software and algorithms powering the self-driving car and license such technology to car manufacturers since Uber has significant expertise in artificial intelligence and robotics.
Currently the technology powering self-driving car is very expensive but their price is expected to decrease exponentially like many other computer technologies. If it cost $50k extra to build a self-driving car, amortizing over 5 years is $10k expense vs the current annual salary of a Uber driver around $50k. This saving can help Uber to recoup the expense to own the vehicle and lower the ride fare for passengers.
One key differentiator Uber can play in developing the technology is to leverage their large network of Uber drivers. To improve the intelligence and reliability of self-driving technologies, extensive road testing and sensor data collection is needed. Companies like Google are building and operating a small fleet of self-driving cars that require full time employees on board. Instead Uber can retrofit existing vehicles with sensors and computer system and offer Uber driver a small stipend to collect data on the street. This crowdsourcing method can be much cheaper than Google’s and enable a larger scale of road testing and data collection, creating an edge over competitors in terms of software and algorithm performance.
With this edge, Uber can also explore licensing self-driving capability to truck and delivery industry. Currently in U.S. the total number of truck and delivery drivers employed are 1.6 million and 800 thousands respectively, a combined market of $90 billion . If Uber can capture 10% of such market with its self-driving technology, that equals to a revenue of $9 billion each year.
 Critical Reasons for Crashes Investigated in the National Motor Vehicle Crash Causation Survey, NHTSA
 Ten ways autonomous driving could redefine the automotive world, McKinsey
 Uber’s self-driving cars have come to Pittsburgh, but that doesn’t mean the driverless era is here, Quartz
 Uber: Changing the Way the WOrld Moves, HBS Case 9-316-101
 Light Truck or Delivery Services Drivers, Bureau of Labor Statistics
 Heavy and Tractor-Trailer Truck Drivers, Bureau of Labor Statistics
Student comments on Uber’s Bet on Self-Driving Car
Peter, thanks for starting this discussion on self-driving cars. While I completely agree with all the benefits you have highlighted, I think self-driving cars pose a deep ethical/moral question that I struggle with. Even if we assume that the cars work perfectly and have no element of human error, these cars are still driving on streets that have humans — passengers in the car, other drivers in cars, pedestrians on the street. Let’s say there is one irresponsible pedestrian crossing the street at the wrong time — if there is a person behind the wheel they will make a human judgement on which way to swing the car. Do they go straight and risk killing this pedestrian? Do they swerve left and bang into the car in the other lane? Or do they swerve right into a wall and risk their own life/life of the other passengers in their car? How will the algorithm be created to prioritize human life? While the pedestrian is at fault and the driver in the other lane is not, if the car hits the pedestrian he/she will most likely die but the other driver will probably be protected with the air bags. Should the algorithm penalize the pedestrian being unsafe or the driver being safe? As we adopt technology we must be keenly aware of its implications. I am personally not sold on self-driving cars just yet as there are too many unanswered questions.
I understand the moral/ethical questions are big concerns of this technology. Many of these questions do not have clear answers even to a human. Different people may answer those differently and the society tend to accept the moral dilemma people face (for example, the court may find a driver killing pedestrian not liable/guilty if he or she is responding to some uncontrollable force in an emergency and his or her own life is in danger). Therefore I feel some of the discussion is a bit unfair towards the technology as we accept the imperfection of human judgement and morality but demand higher moral standard for technology.
This also reminded me some irkness discussed in class. Certain price/tax strategy usually benefits some while harm others. What matters most to me is that the overall social equity or fairness is improved. In the case of self-driving car, if it can indeed reduce the 90% of fatality by taking dangerous drivers off the road and offering more responsive driving using different sensors, I think the net gain is significant enough even if the technology occasionally made mistakes in some tough moral dilemmas.
I’d love to continue the discussion offline and feel free to reach out to me.
I think some of the ethical questions that come up with regard to self driving cars are actually less impactful than we may think. For example in the case brought up here, where the car needs to decide whether to hit a misbehaving pedestrian or innocent car, the algorithm probably won’t be making any ethical judgments whatsoever. It will just be trying to minimize a collision. It’s actually pretty hard to bake ethical logic into the algorithm. Moreover, I’d argue that humans aren’t particularly good at this either. If you’re in a split second decision you’re probably not thinking about what the most ethical course of action is. Your reflexes are probably taking over and just trying to minimize the chance of a crash
Fascinating issue, Peter. It is amazing (but not surprising really) how many traffic accidents are caused by human error. But even if driverless cars can reduce this number significantly, I still think that society will be very reluctant to adopt this technology. The idea of giving up control, even if we know it would be statistically safer, is still a significant hurdle for people to overcome. Does Uber intend to own a fleet of self driving cars, or do they merely intend to provide a platform to connect cars to passengers like they do now?
I added the “Future Business Model” section which explains why Uber have to own a fleet of self driving cars since fewer people will own cars in future. Many industry experts believe autonomous vehicles will begin with truck and delivery industry as business has strong interest to save cost and no passenger is involved or giving up control. Once the society becomes more comfortable with many autonomous truck and delivery vehicles on the road, it will become easier for general public to accept the self-driving cars.
I am glad to see that the self driving car is right around the corner! I hate to drive and have joked for years that by the time I have to drive regularly I will be able to rely on the driverless car. I thought your point regarding reduction of personal ownership was quite interesting. Previously, I had thought each person would own their own driverless car, but now I see the potential for the combination of ride sharing and driverless cars to result in less of a need to own. I do think it will be a while until we get to this state for a few reasons (1) because of the concerns mentioned in comments above regarding trusting a machine vs. a human and (2) I’m not sure how this scales in non-urban areas given variable demand (3) I think daily car commuters will be reluctant to share any time soon. Do you think this will eventually change all ground transportation or do you think it will remain a niche opportunity for years to come?
This post smells dangerous. While I have always been excited about self-driving car and what the world will look like when that comes to life, recently I can’t help to think about the “negative” impact it may bring. This is essentially another machine replacing human story. Bus drivers, taxi drivers, delivery man will all lose their jobs. Cyber security is another concern. If the control system is hacked by terrorists, I can’t imagine what will happen to all the driver-less cars on the road. Will they crash into each other?
I beg you not to think this way. This type of thought keeps humanity from progressing in a logical way. Just because a human has done a job in the past does not mean that a human should continue to do a job into the future. Yes, there does exist a far away future where humans are not needed for a majority of the jobs they do today. Technology has created value far beyond our financial instruments and will continue to do so into the future, if we allow it. This is not a future we should be afraid of, but rather one we should work to shape to be fair and equitable for all.
Cyber security is no different than a road side IED – a constant threat that we work diligently to deal with. Those with evil intentions are present whether or not we adopt this technology…
I have to agree with Denton. In terms of displacing labor, it is a tale as old as time for people to be losing their jobs to machines and fleeing higher up the value chain. Regarding security concerns: a popular theory in the industry to is assume that those with malicious intent will always have the upper hand: a terrorist only needs to find the backdoor once, cyber security only needs to fail once to fail entirely. That as a given, should we simply stop developing past the point where our technology has the capability to cause mass loss of life and/or mutually assured global destruction? This I am not as clear on. N_N brings up a relevant secondary point: when should we begin to fold in the scale and magnitude of our technology into the ethical discussion of should we do this, or shouldn’t we? Currently we are already seeing many of our greatest minds wrestle with this topic in the artificial intelligence community. No clear answers yet.
Peter, thank you for your post. I think the possibilities of driver-less cars (both in safety and productivity) are fascinating. Was wondering if you saw anything about the regulations that may be coming with such new technologies. It seems to me that the benefit of self driving cars is reduced if cars with drivers are still allowed on the road. Do you see a future in which driven cars are eliminated/banned? What would this mean for people who enjoy driving as a leisure or sport?
Do people still ride house for commuting after automobile is invented? Yes, but for a short period of time (maybe 20 years, from 1890 to 1910) and eventually it becomes an expensive leisure and sport. Same applies to manual vs auto transmission. Most racing cars still have manual but it is becoming a luxury for fewer people.
If self-driving can be intelligent enough to understand human driver perfectly while be more responsive of the surrounding environment, it can coexist with human driver and be more efficient and safer. While I do agree the benefit of self-driving car will be limited if cars with drivers are still on the road, the benefit can be big enough to incentivize people choosing self-driving car. I don’t think we need to ban car with drivers. The saving on insurance, convenience and fuel should gradually make cars with driver obsolete.
Thanks for the interesting post. Like others who have commented, I too am curious about the dangers of self-driving cars. It is clearly expected that there will be fatalities related to self-driving cars ( http://www.nytimes.com/2016/07/01/business/self-driving-tesla-fatal-crash-investigation.html?_r=1 ). Tesla’s response to the May 7th fatality was troubling in that it mainly discussed statistics, and saved the final sentence for condolences to the family of the driver. ( https://www.tesla.com/blog/tragic-loss?redirect=no ) As the NYT noted, the crash casts doubts on whether autonomous vehicles can make split-second driving decisions on highways. Do you find these doubts to be justified? Is Tesla at fault, the driver, or the components manufacturers such as Mobileye who created the technology behind ADAS (advanced driver assistance systems)?
You did an awesome job bringing up Tesla technology and recent fatality. To be fair, Tesla’s auto-pilot feature is rated as level 2 assisted driving and is never intended to be used in fully-autonomous mode (level 4). It is partially due to Tesla’s mis-advertising to make people believe it can self drive and people rely on it recklessly. Tesla’s Mobileye solution costs around $1,000 vs Google and Uber’s solution costs around $100,000 (due to the expensive Lidar) so it is no way representing what the truly autonomous vehicle’s capabilities. That being said, even with Tesla’s lousy solution, it already achieved a fatality rate lower than current US driver. In US, on average there is one fatality for every 100 million miles and Tesla already accumulated around 200 million auto-pilot miles with one fatality. This number is not statistically significant yet but we can only hope the next gen technology will be much safer.
Regarding to highway safety, it is actually found that driving on highway is much safer and simpler than driving local. Google started road test on highway first and then move to local. It is more predictable environment with less variance. The cognitive load is also simpler on highway than local as objects and intentions are easier to understand. So industry in general does not worry much about highway compared with local environment for self-driving cars.
Thanks Pete, regarding the point on increased worker productivity from being able to work while being driven, Im curious as to whether this is being captured in economic statistics. With recent declines in productivity, there’s a popular view that economic statistics are not being measure correctly. This applies to uber in two ways, the first I already mentioned before and the second regarding drivers that go on uber for supplemental income.
Interesting post! As many have mentioned, while there are too many car accidents due to human error each year, I think the public will react more extremely and negatively if there are any car accidents due to driver-less cars. The lack of control when it comes to driving is a new concept for consumers; therefore, the spotlight on the number of accidents due to driver-less cars will be so extreme, that any one accident will be known amongst the public. Moreover, as we have seen with computers and all technology, there are always new and updated models coming out (i.e. iPhone – iPhone 7). Each model update costs money, so what will be the impact of costs on the driver-less cars? Will Uber have to update cars each year in order to stay safe?
Actually Tesla is already updating their software regularly to push for new navigation and auto-pilot feature. It is just like iPhone software update and the frequency is much higher than once a year, roughly every month they will drop a new build. Obviously we expect Uber or any other self-driving car company will have this OTA(Over the air) upgrade capability.
While I think the technology is fascinating, I also wonder about the economic impacts it may have. As you noted, car ownership will become burdensome in relation to use of self driving vehicles, which may punish those living in rural areas that don’t have the population density to support a fleet of autonomous vehicles. Additionally, autonomous vehicles may eliminate a large number of jobs in the transportation and freight industries, leading to further economic impacts and perhaps more exaggerated income inequality as low-income jobs are removed from the economy.
Very interesting article, Peter! I too am a bit worried about the widespread use of self-driving cars. While I recognize that over 90% of car collisions are caused by human error and this removes the human error element, I am very concerned about cyber security threats. Terrorists may have the ability to hack the self-driving vehicle and use it remotely. Or there is even the potential for cyber terrorism through a large-scale attack of cars in a particular country. I’m sure much research has gone into ensuring these vehicles are safe and less likely to be hacked. Have you come across this in your research?
Thank you for bringing up the topic of cyber-security. It is indeed one of the biggest concerns for IoT in general. One idea in security is about attack surface, total sum of the vulnerabilities in a given computing device or network that are accessible to a hacker. The attack surface definitely increased over the past decades as we are becoming more mobile connected and more data are stored and accessible from the internet. Although both public and private sectors are putting lots of efforts on cyber security, I feel that the growth of IoT and other connected devices is faster than the development of cyber security.
Short term approach would be increase the investment or regulation around cyber security, making it a mandate for many critical services. Long term I expect industry and research come up with intrinsically more secure mechanism to decrease the vulnerabilities, e.g. block-chain and quantum cryptography.
Uber’s driverless car sounds promising but I do believe the real at-scale adoption of the autonomous cars will be in the transportation industry, especially in the long-haul trucking industry. Certain state highway systems already have dedicated truck routes/lanes so the infrastructure is more ready for such technology.
Thanks for opening a fascinating discussion about self-driving cars – very hot topic! Indeed, over the last years, there have been several stories in the news about the development of self-driving cars. Several states have debated legislation re. authorizing self-driving cars. However, only a few number of states have enacted such laws. Do you think such trend is reversible? On the one hand, there is a legal vagueness that lets one assume that autonomous vehicles may technically be allowed to operate over the roads – if a human being sits behind the wheel. On the other hand, laws require that a “human wheel” operates the care. Which direction will laws/regulators embrace?
Very interesting, Peter! I really like your idea on Uber partnering with car manufacturer to create driverless car services. I certainly agree with you that this model would allow each party to bring in their expertise and competitive advantage – navigation technology and enormous rider base from Uber coupled with driverless car technology at low cost from the manufacturers. I however doubt the practicality of leverage Uber’s network of drivers to help with road testing and sensor data collection. I believe that most drivers would be hesitant to collaborating in the development of driverless car technology that could eventually steal jobs from them.
I like your concern in the end that driver may be hesitant to collaborate in the development of driver-less car. However I feel most people do not think too far beyond their direct benefit. Earning some money now vs potentially losing job years in future, I am sure enough people will be interested if the price is right.
What concerns me a bit about self driving cars is the potential negative impact on the economy for a couple of reasons:
1. Putting the millions of drivers out of work could lead to a serious economic downturn and some serious civil unrest.
2. If cars are fully utilized instead of the current 5 percent utilization then obviously a lot less cars would need to be produced. This of course would have some serious negative implications for the auto industry. A lot of factory workers would lose their jobs.
With the potential devastating effect to the economy I wonder if politicians will be pressured to regulate away the driverless car.
Very interesting point that Uber’s current competitive advantage in self-driving cars lies in the size of their existing network, whereby they should be able to train algorithms using their current drivers. I have been thinking a lot about which company is most likely to emerge as the early leader in SDCs–Uber’s strength is not only the point that you bring up regarding their network, but also that they have a proven track record of taking on regulators and winning. Google, however, will likely be able to offer the lowest cost service over the long-term, since they should be able to subsidize fares by displaying ads inside the vehicles.
I’m not as concerned about the ethical issue of prioritizing patient vs. pedestrian safety. I believe the issue is likely to be a corner-case where there isn’t a very clearly identifiable “safest move”. However, I believe to the extent that this logic does need to be built into the algorithms, the algorithm will have to essentially always prioritize the safety of the passenger. Otherwise, lack of passenger trust in the system could erode the viability of SDCs as a service.
Peter, thanks for posting about self-driving cars. As someone that lost family members in a car accident, I am excited about any technology that can decrease human involvement in car driving. Many others posted about the implications that self-driving cars may have on the economy and job market, but I think there is a big question mark over how regulators will react to self-driving cars. I think governments have proven themselves sluggish to address technological changes and the introduction of self-driving cars, as you mentioned in your post, is the first drastic change in the automobile industry that will impose huge challenges to policy makers.
The advent of driverless car technology is exciting – as it has the potential to disrupt a substantial portion of human behaviors and consumer patterns. While I think ultimately we will head in that direction, I would caution about the immediacy of its impact. Google engineers tell a story about a car sitting in an intersection while a lady dressed as a witch chases a duck around the intersection with a broom tracing the shape of a figure 8. The autonomous car will just sit in the intersection, blocking traffic, unable to process the situation. Most humans would just observe the occurrence and pull around the duck and witch, but the computer, as of now, does not have that ability.
While utopia seems great on paper – I believe we are a long way from achieving that.
Fascinating article, Peter. While it all does sound very good, I have to agree with both Alex H and Alex in terms of my concerns about this technology. The impact and implications of these self-driving cars goes much further than just saving on car ownership, reducing insurance premiums, and eliminating traffic. There are over 3M workers in the auto manufacturing and sales industry . If our utilization of cars increases as you predict, that will significantly lower the demand for those workers, possibly leaving millions without jobs. As a consumer, the thought of self-driving cars sounds great but as a member of society, I’m concerned.
I think the goal of self-driving car is not to replace human but to improve the overall efficiency of the society. We used to have 90% of population working as farmer and now that number is less than 1% in U.S. while the society’s productivity and quality of life is much improved. We also witness in the early 20th centuries many home appliances (such as washing machines) saved hours of time for housewives so many of them lost jobs as full-time housewives or cloth washer for other families to become working professionally.
Mid-18th to early-19th century saw industrialization–the process disrupted and replaced low-skilled labor with engineering might. With the development of artificial intelligence, more low-skilled labor will be inevitably replaced. In fact, bots now crawl data and generate financial news. For example, robo advisers are now helping us make investment decisions. Uber replacing its “partners” with robots is a matter of time. Whilst some posters have expressed concerns over ethics, we mustn’t underplay the number of lives that would be saved because of human error. Furthermore, companies are now combating these challenges by coming up with creative solutions. U.S. Patent No. 9,429,947 describes a mechanism allowing those outside the self-driving car to interact with the path of the car.
Peter – Thanks for the very interesting post on self-driving cars. I admit, I am a big believer in the potential impact for good that self-driving cars can/will have in the future. I appreciate that some may find the thought of giving up the role of driving their car to a computer as scary–I think in 50 years we’ll look back and marvel that we ever let humans, prone to bouts of sleepiness, intoxication, distractions, etc., to drive cars! I found your point regarding leveraging the Uber drivers to gather data in order to improve the reliability of self-driving cars to be interesting. While Uber does have a fleet of employees at its disposable, I do wonder if an otherwise self-interested Uber driver would be interested in helping to perfect the very system that will put him or her out of a job. I suspect only those who feel the adoption of self-driving cars is already inevitable and want to generate any economic value off of it in the meantime would do such a thing. Still, very interesting to consider how self-driving cars will impact Uber’s value proposition and cost structure, as we discussed in class last week. Thanks for the great post!
Thanks for this post Peter. I find the possibility of self driving cars very fascinating because of some of the benefits you have mentioned in your post such as increased productivity, better for the environment, cost savings, could save lives (with some researchers estimating that up to 300 lives a year could be saved if autonomous cars are widely adopted – http://www.theatlantic.com/technology/archive/2015/09/self-driving-cars-could-save-300000-lives-per-decade-in-america/407956/) etc. But it can also be a contentious issue with some interesting ethical dilemmas. For example, this Business Insider post – http://www.businessinsider.com/new-study-reveals-people-want-self-driving-cars-to-sacrifice-passengers-during-emergencies-2016-6 – describes a recent survey in which respondents had conflicting thoughts on what an autonomous cars should be designed to do in an emergency, with many preferring that the car protect them at all cost during an emergency, even if it meant harming pedestrians or other drivers. It will be interesting to see how such ethical issues are reconciled with the design of self-driving cars.
Interesting! Looking at Boston Uber Pool pricing I am a bit concerned that even if they replace the drivers with autonomous cars, it will be still far from breakeven:) One open question for me here is who will own the fleet – Car manufacturers? Uber? Private users?
Currently in the Uber ride fare, I believe 60~80% is due to driver salary (depending on the country). Therefore self-driving has the potential to lower the fare to 1/3 or even 1/5 of current level. My argument is that at such price, no one wanna own private car any more therefore Uber will have to buy their own fleet. The operating cost of maintaining such fleet will be much lower than what we currently envision too, for example the cars can automatically park and recharge during the off-peak hour and self-diagnose many issues it detect.