The Self-Driving Car - Will we Trust This Technology and How Will It Impact Our Future?
What is all of the sudden hype about the driver-less car? Within the next five years, and even as early as the end of 2017, Google's parent company Alphabet's Waymo, Uber, Tesla, Mercedes-Benz, Samsung, Nissan, Jaguar Land-Rover, Toyota, Ford, GM, and numerous autonomous vehicle startups have all ambitiously pledged to release their version of the self-driving car for public consumption. These companies are in a bitterly vicious competition to be number one in this new paradigm, and are forming various corporate alliances to gain competitive advantage. The technologies and types of products vary widely in their undertakings and technology, ranging from an autonomous taxi service, vehicle subscription service, and finally to consumer owned and operated auto-piloted cars. However, they each require us humans to move to a new level of trust and allow an artificial intelligence to take the wheel. It's truly difficult to imagine how this will all play out in the end, as there are many opposing forces operating at the same time. Cars, car ownership, and the means of transportation have so much more significance to so many rather than simply getting from point A to point B. Simultaneously, the consumer side of the auto industry is continuing to change, with the shift in focus on the environment and renewable energy. Hence, finding the sweet spot to fit this particular technology in a constantly moving target is a challenge. One thing is for sure, when the trends start moving, no company wants to be left behind.
Public Reaction
It's not really clear how the general public will react to self-driving cars. Some trials have yielded positive results, like Uber in Pittsburgh, where people reported receiving uneventful rides in a real city - achieving the goal of a normal experience in the car. However, in these scenarios, there were backup safety drivers ready to fully assume control, even if they didn't need to take over that often, which can give it's own measure of security. Not to mention the ride was free, which offers its own reward. Waymo has now offered fully driverless trips to non-employee passengers, including a blind gentleman from Austin Texas who hasn't "driven" since he lost his eyesight. These trips have also been successful, both on and off the test track, with reports that the car behaves cautiously and respects bicyclists and pedestrians. The aforementioned gentleman has reported the newfound freedom that being able to be alone in a car gives him; a sentiment expected to be shared by many others in a similar situation. However, the silent majority will be hard to predict in this case. Skimming the comments section of any article on self-driving cars reveals both the proponents and the skeptics. The proponents report increases in safety by eliminating human driving errors and providing traffic relief. The skeptics state that these cars are an "abomination" or "death-traps", which will interfere with human drivers, or cause more accidents on their own. Another major issue to acknowledge will be the necessary shift in psychology of drivers. The car represents an extension of our identity, status and control. The self-driving cars have to be positioned as stylish, personally relatable and viable alternatives, advantageous enough to compel people to be willing to let go of a certain portion of themselves and their ability to directly control their mobility. Importantly, the cost and level of convenience will definitely be a factor, as people will soon be deciding between mass transit, bicycles, regular or autonomous personal vehicles, and regular or autonomous taxi services to get around. Finally, the autonomous vehicles must be convincing as a truly safer alternative to human intelligence behind the wheel, for drivers to trust and willingly participate in this transformation. Otherwise, in a potential dystopian outcome, there could be a major battle on the personal transportation front, and the forces with the greater power will determine the direction the future takes.
Benefits and Concerns
There are several potential benefits to self driving cars, however the number one touted benefit is in the important realm of safety. Here most articles will cite the statistics of automobile accidents, death and disability tolls for drivers, passengers, bicyclists and pedestrians due to human error. This is caused overwhelmingly by distracted or impaired driving or poor judgement. The idea that automation will be safer than human drivers, stems from the fact that a machine is not subject to distractions or human impairments, hence it will not be prone to make errors and cause accidents. Furthermore, ideally the cars will be able to communicate with one another in real time, without ego or road rage, to avoid ambiguity and collisions.
Yet, anyone who has ever used technology knows that it is not perfect, and any glitch while moving at 35 mph or greater can be deadly. Although Waymo, for example, has logged eight years and millions of miles in a combination of real life and virtual trainings, and the computer has much more information about the environment at any given time than a human being ever does while driving, the computer requires that much more info in order to avoid perception errors that a human would not make. If the car is navigating with Google Maps, we already know the potential errors that can occur just considering navigation alone. This does not take into account computer recognition errors such as occurred in the fatal Tesla autopilot crash where the sensors mistook the white side panel of the tractor trailer for the skyline. So as a rider of an automated vehicle, one is essentially hoping that all of the scenarios which the car has been trained in, are similar to the one in which they are currently in, otherwise the automated response may not be what we think it should be. The intelligence behind the automation may not be subject to human flaws and influences, but it's also not as smart as humans either, and lacking in value judgements - both of which makes the prospect of relying upon it frightening. Although tech companies are constantly improving their ability to reduce these occurrences by enhancing their models, and even discussing ways to train the car to behave ethically in an emergency, we may be simply trading human error for computer error in causing fatal crashes. Which scenario would you prefer...or more directly, which driver (human or machine) would you bet on to ensure your safety?
There are benefits of self-driving cars in reducing traffic congestion as many of the issues related to "stop-and-go" traffic are due to human selfishness, impatience, and poor driving maneuvers. Ideally machine driven cars would function in a smooth coordinated fashion with speed reductions, better inter-car spacing and more permissive maneuvers. The headache of driving in or experiencing the nauseating "stop-and-go" will be eliminated with this new paradigm. An added bonus would be improved fuel economy, as more efficient motion would reduce gas consumption. However, this would be less and less important as we move towards electric and other renewable energy cars.
Nonetheless, there are three caveats to this traffic ameliorating benefit. Firstly, it would require that the majority of cars on the road are self-driving cars. Having human drivers sharing the road with self driving cars would interfere with the possibility for this automated car synchrony. Having a majority of autonomous vehicles on the road would require near total participation by consumers which may be difficult to achieve. Secondly, even if the cars drive in a coordinated and synergistic manner, pedestrians, bicyclists, and random objects may not follow the rules. In heavily populated locales, such as NYC, pedestrian traffic and other randomly moving or positioned obstacles can still negatively influence the cars' ability to maneuver smoothly and affect overall traffic motion. Finally, with a large amount of cars traveling and more permissive driving style, it will likely take much longer to reach a destination. A human driver may take shortcuts, or not allow certain maneuvers from other drivers in order to reach their destination within a set time. A computer will follow the rules and go slowly; with heavy traffic, this could mean seemingly unbearable delays. Will us ruthlessly busy type A personalities be willing to sit still in a slow moving "dumb" smart-vehicle for as long as it takes to reach our destinations?
Perhaps the time-pressured and goal driven may accept the above challenge as a trade-off to the time the self-driving car can give them to do other things. For those of us who tend to drive ourselves to various destinations, having your own car shoulder some of the effort may be a welcome benefit. The benefit of saving time in actual driving, sitting in traffic, and looking for parking, is not to be underestimated. This same exact benefit can be achieved with hailing a regular taxi instead of driving. Nonetheless, it's hard to ignore the obvious joy of stepping outside to have your own car greet you, drop you to your destination and then go park itself without your input. Many have gone even further to reimagine the car as a sort of living room or lounge on wheels, and designed changes in seating to better suit a conference or party rather than a travel vehicle.
There is another group of people who would directly benefit from self-driving cars, and they are those who have lost their ability to drive in the conventional fashion. Those who are elderly or disabled may have had a significant life change causing them to have to give up their freedom; it impacts them considerably. Having the ability to move about without needing to rely on others is a major advantage to self-driving car technology. Click the link for a further discussion of reasons to embrace self driving cars.
Challenges
Nonetheless, as wonderful as some of the benefits seem above there are also challenges unique to autonomous cars that seriously rival the benefits. These questions elucidate the very tricky territory which must be navigated to implement self-driving cars into the mainstream of society.
How will they be regulated? How can we ensure a reasonable measure of safety with varied companies and technologies emerging to build their products in unique and proprietary ways? Generally cities and jurisdictions solve this problem with laws and rules, and federal governing bodies such as the NHTSA. However, these have all existed on the model of a car operated by a human being. Trying to ensure safety in a technological patchwork of corporate approaches to self-driving may require an additional technologically dedicated governing body to evaluate each specific iteration of autonomous vehicle. The NHTSA has currently established standards to evaluate the levels of automation each vehicle has in order to begin to categorize these technologies. Other issues, such as privacy may be a concern now with these new technologies. Congress passed a bill to help address this problem in Sept. 2017.
Who is liable for what? If your self driving car gets into an accident...are you liable or the auto-maker? Is the self-driving technology at fault or the actual automaker (because they may not be one and the same). How will insurance companies be able to sort out liabilities? Apparently the AI in Google's self driving car is considered to be a legal driver. How will insurance companies be able to determine risks to determine premium pricing when dealing with the unknown of automated transport in a changing environment? What will it be like, and will it be reasonable at all for the consumer to have to pay for insurance to cover potential scenarios over which they have no direct control anymore?
How will the car make life or death decisions? What reasonable and ethical way will there be to determine what to tell the computer to do in a split decision which could mean life or death? You would hope that those situations wouldn't occur often with autonomous cars, however you can't eliminate the realities of life. Any algorithm would seem to somehow have to make life value decisions which are obviously fraught with conflict. A human might decide to sacrifice themselves, but would they expect their car to sacrifice them? Would consumers even be able to gain knowledge of which life-or-death algorithms are operating in the car they are riding in?
What if you encounter a situation that requires you to change course, during travel, can you update the machine in real time with your new intentions? Sometimes life doesn't go according to plan. Sometimes personal events, weather events, etc. cause you to have to change your course. How quickly and properly will the machine respond? Waymo has a button you can press to contact a human to remotely control the car, however, does this make you feel secure in a decision which must be made quickly? How each company solves this problem will determine whether people feel the tech is more of a hindrance than a convenience. Ultimately it comes down to how can a person continue to maintain full control of their transportation when in a self-driving vehicle? As a passenger in mass transit, such as a plane or train, each person understands they have given up a measure of control. However trying to echo this phenomenon in the personal transportation arena, may be asking for more control than many people are willing to let go of. Especially when the control is not being given to a human being but to a computer algorithm which they don't understand and can't communicate with.
How much will it cost? Having high end technology in your car which allows it to drive itself will undoubtedly increase the costs of personal ownership. The luxury vehicle consumers or automakers may have no issue absorbing the extra costs if the benefits are substantial. However for the autonomous car to have widespread usefulness, it would have to be included in regular production vehicles and the extra cost could not be prohibitive. Otherwise people will opt out of this feature on cost alone. A self-driving taxi service may get away with lower costs due to not having human drivers, however will this service encourage people to give up the convenience of car ownership? Riders will still have to pay each time they use the service and how would that compare to owning an electric car and just getting in to drive when it is needed with little cost? It may be hard to predict which will appear to be more cost effective, but people are likely to gravitate in that direction. The other cost which must be considered is the cost to cities in now having to design infrastructure to support the self-driving cars. Will taxpayers now be forced to pay extra taxes to build all the millions of new road signs and street markings which must be established for these new vehicles?
How to prevent cyber attacks and privacy breaches? This is one of the terrifying questions of self-driving cars - are they hackable, and could they be weaponized? There is no doubt however, that human driven cars can be used for terror as they have been in several incidents, most recently in NYC. Yet, can targeted auto cyberattacks or large scale hacks be possible in these autonomous vehicles and used for terror, kidnapping, or any number of ill-purposes? Additionally, autonomous vehicles operating on GPS will have position and location constantly recorded and available; one would wonder who could gain access to such information. The problem with these potential attacks is that they can be done remotely, AI assisted, and anonymously, making it even harder to address, control and prevent, as compared to the usual risks associated with human piloted auto attacks.
How will people agree to accept self-driving cars? With the above benefits and concerns, it will be difficult to reach a consensus as to how to proceed. Obviously those with the most to gain will want to move forward (i.e. companies which have invested billions, and individuals unable to drive), while those with much to lose, individuals with lower financial means who could end up paying more, or anyone who desires greater control over their personal transportation will be less willing to welcome the autonomous vehicle. The appearance and likability of self-driving cars may also play a role in their selection on the market. Perhaps people will be fine riding in a sleek and beautiful Tesla, but see the rounded, minivan, carpool-efficient, family vehicles as not "sexy" enough, or just not representative enough of themselves, to warrant a purchase. If the production cars are not cost effective, electric or renewable energy derived, they may also take a hit in terms of sales for a more efficiency and conscious minded market. Not to forget, the entire prospect of fully autonomous car takes away the pleasure of driving for those car enthusiasts among us. Finally the questions of autonomous vehicle safety are truly unknown, although theorized to be safer, as compared to the familiar risks of human controlled driving. Despite these varied opinions, most of the benefits of self driving cars will come from a majority, if not all cars following this model. However, for the supposed "average" consumer, there seems to be no sine qua non of a driverless car that will make them abandon their personal vehicle and convert to a taxi service, or try to invest in one of these autonomous cars. So will an attempt be made by automakers (via limiting choices) and government (via rules and regulations) to force people to accept this mode of transportation so that the potential benefits may be realized?
Click the link in this sentence to read about the disadvantages of self driving cars.
Impact on the Future of Transportation
In conclusion, the impact of driver-less cars on the future of transportation is an extremely complex determination to make as there are so many factors involved. This is notwithstanding the fact that achieving a safe self-driving car that can adequately mimic what we are currently used to in various situations (although, ideally without the downsides) is extremely difficult. "It turns out that driving is a very complex activity. In fact, it may be the most complex activity that most adults on the planet engage in." Raj Rajkumar an early self-driving car engineer of Carnegie-Mellon states in All Things Considered for NPR. And for this reason self-driving cars will have to spark a brand new system of functioning in transportation than the one we are used to in order to progress. Like many futuristic endeavors, it will be hard to predict the outcome of this proposed direction as it can't be compared to our previous experiences. We will have to watch as this momentous endeavor unfolds over time to grand or catastrophic consequence. Did you realize that a potential transportation revolution would be coming this soon? We'd love to hear your opinion. Feel free to leave a comment here or on Twitter and check out our other blog posts.