Driverless cars pros and cons - Are self-driving cars worth the data sharing price?

We outline the top ten pros and cons of self-driving. Join our discussion on the emerging problems associated to data collection and sharing by manufacturers of driverless vehicles. Find out more about the regulatory plans of NHTSA and the attitude of the industry towards it. 

Debate created by Paul McNeil 

Are self-driving cars worth the data sharing price?

Self-driving cars, also known as autonomous or driverless cars, are forecasted to radically change the very nature of commuting and transportation. However, these vehicles will be filled with huge amounts of information, turning them into high valuable targets for manufacturers to begin with, the government, and of course, the hacker community. This opens a whole new world of threats and brings this question to mind: What are the pros and cons of having unmanned cars roaming in our streets? Is the entire concept of autonomous cars worth the price of the information poised to be shared about our commuting habits? Is this yet another sector of our privacy to be breached and lost forever?

Self-Driving Cars Pros and Cons

As with every new innovation entering the market, evaluating the advantages and disadvantages is a must first step. Here you can find 10 pros and cons associated to the introduction of driverless car technology

Pros

1. Increasing safety by providing blind-spot assistance to the driver. There are less opportunities for human distractions and excessive speed which cause most of the accidents. Drunk driving incidents would decrease
2. Remote car ignition, A/C, heater and locking doors made possible through smart phone apps
3. Artificial Intelligence (IA) allowing warnings to alert drivers before straying across a solid line on the road
4. Vehicle location apps indicate if a car has gone beyond designated area, for example, by your teenage son
5. Parallel parking made easy through automation. Cars could drop off the passengers even before having finished parking
6. As computers take over the responsability of driving, traffic conditions and congestion would significantly improve
7. Drivers could devote their time and attention to other things, such as reading, working or chatting with passengers
8. Police could spend less time monitoring traffic and focus on other more important issues
9. Stimulating effect in the automotive industry
10. Disable individuals will not have to rely on public transportation

Cons

1. Coming at a high cost while being imperfect at first
2. Driverless cars are not completely safe and there have been a number of accidents involving them
3. Limited abilities in self-driving cars forecasted way into 2030s and middle class families won’t enjoy such cars until at least 2040s
4. An ethical dilemma emerges from computer using algorithms to replace human choices while driving, in particular when it comes to reactions to incidents (watch video below)
5. By relying on this technology, people may gradually lose the skills to drive cars which they may need in countries or areas where this technology is not operational
6. Not everyone prefers self-driving or will adopt it. Mixed traffic can render even further problems
7. The popularization of self-driving vehicles could undermine the efforts in big cities to promote public transport to reduce traffic and pollution
8. Disruptive effect in the automotive industry which could make many people lose their jobs
9. The computers monitoring self-driving cars could be hacked and their information leaked. Hackers could even take control of self-driving cars. Therefore both privacy and security are would be at risk
10. Data collection and data sharing requirements to operate vehicles and reduce accidents may create further problems (see next section)

Watch this TedEd video on the ethical dilemma of self-driving cars

Crash Data & Quarrelling Companies

Driverless car are packed with state-of-the-art technology are considered information goldmines for an increasing number stakeholders, including car manufacturers, governments and hackers. In response, the American National Highway Traffic Safety Administration, NHTSA, which is part of the Department of Transportation is first ensuring its own, most probably exclusive access to this rich data bank. In the recently released Federal Autonomous Vehicles Policy, the NHTSA states its intention to oversee the development, regulation and policing of self-driving cars inside the United States. The agency is hoping to expand the guidelines’ reach through collaborating efforts involving America’s only two land neighbors, Canada and Mexico. Autonomous vehicle companies will be pushed by these federal guidelines to begin sharing data regarding their failures with even rival companies, and also the government. This will be considered a serious shift already faced with stiff resistance from tech and auto companies that enjoy significant influence inside their respective governments.

In the case of self-driving cars crashing, the resulting big data should be retrieved by the NHTSA and the car manufacturer without any bias, all to facilitate crash reconstruction and the necessary analysis. The least is that vehicles should record all crash-relevant data and system performance, to allow engineers reconstruct the event circumstances. This information will then be sent to federal regulators and original manufacturers. Furthermore, manufacturers will be held accountable for having their customers completely understand how the distribution of the crash data will be carried out.

Under what method will companies, already in tight competition to be the first to mass-produce self-driving cars, will share that information remains an issue at large. However, a certain fact is that tech companies developing driverless cars are not too pleased about the entire concept. Data sharing has even been described as the sharing details with the devil, which has become a sticking point for all involved private companies. This is especially true in regarding to autonomous vehicles.

There is a real chance that the auto vehicle industry will begin pushing against any requirement aimed at making data sharing mandatory. For instance, despite praising the NHTSA’s effort to advance the mobility future, Ford refused to make any remarks about where the company stands on the sheer concept of sharing more information with government agencies, let alone outside groups. The fact is this concept is asking companies to share their data with their competitors, information that is considered confidential. Despite all this, these companies do agree that the customers’ safety always comes first. It has become quite obvious that time is needed for companies and official regulators to determine the fine balance between sharing data for the sake of enhancing safety measures, and protecting a company’s rights to its own data and what is considered the property of manufacturers and innovators.

The NHTSA understands very well that reaching a point where all manufacturers are convinced to sharing their crash data is quite a challenge. The agency is weighing various mechanisms to share data that will maintain an anonymous nature to data and actually avoid any complaints over antitrust issues. Mechanisms need to be established to share data, while it is also an acknowledged fact that soon to be shared data elements will be in need.

Open source is not a concept that car companies are familiar with. Tesla placing its patents available in 2014 to any and all its competitors seeking to benefit from them was one exception that made serious headlines. However, Tesla, the famous electric cars manufacturer, still remains very protective of its driving data that is the power behind the autonomy features of its Autopilot technology. All said, Tesla was eventually forced to provide the NHTSA-requested data logs regarding the May 7 fatal crash of a Model S. The agency also filed a follow-up request that was a bit modified.

The Potential

The major companies involved in the autonomous vehicle industry, such as Google, Audi, Ford, Nissan, Delphi, Tesla, BMW and NVidia,  all agree there are huge stores of data gathered by the thousands of cars roaming the streets under their name. Furthermore, all this data used to make these systems learn –data regarding the methods used by autonomous vehicles to respond to actual impact events and challenging circumstances– may become highly valuable when intending to manufacture a very adaptable and robust self-driving car. If a ridesharing company like Uber can produce a vehicle able to handle a variety of road conditions more difficult than that of its competitors, this will in the end mean better margins and lower fees for customers.

A Window

A solution that has been mentioned so far demands that companies provide their data to a third-party aggregator. This is, of course, another aspect of the overall concept that tech companies will most definitely seek to exclude from getting a hold of their data.

All in all, however, involving a third-party to collect the data separate from the NHTSA may actually provide the benefit of protecting sensitive data from others. If the NHTSA plays the role of a clearinghouse, the information being generated by other companies will become targets of requests made by the public. This will allow the ordinary people and journalists to take a look at the information. Uber has already been involved in a battle with each state in the U.S. regarding these very matters, showing high sensitivity about keeping data on its drop-offs and pick-ups a secret from the outside world. Moreover, making data collection and data sharing requirements at the federal level will permit companies to avoid the variance of state-by-state policies. Working with a number of agencies at the state and local level is something most companies want to elude.

The NHTSA is seeking a widespread data sharing culture far beyond just crash data. The agency is also calling for companies to begin sharing data after experiencing a hack attack. For instance, after a Tesla car was hacked by Chinese individuals that went as far as claiming to take control of the car’s breaks and even the mirrors. The industry must share data on cybersecurity issues in order to guarantee all the data at their disposal. Each member of the industry should not need to undergo the same mistake or experience similar cyber vulnerabilities to also learn from.

In the meantime, the NHTSA is also seeking to get its hands on data showing various vulnerabilities in autonomous driving. And this may actually turn out to be a beneficial factor for the general public.

The Vague Aspect

Despite the fact that Washington has placed forwards major reservations on the design of self-driving cars, safety features and data sharing mechanisms, the requirements raised for cybersecurity issues can only be described as vague. NHTSA policy demands manufacturers are required to impose measures aimed at protecting data related to losing data or disclosing it under unauthorized terms. However, in the end the companies are left to determine what harm might come through such practice. A manufacturer found introducing security vulnerabilities should be held responsible, especially if such loopholes allows hackers to literally take over and force a car to crash. However, imagining scenarios where companies may attempt to deny any responsibility is not hard at all. For example, what happens if a local abuser follows his victim by taking advantage of hacked geolocation data from a self-driving vehicle, and actually uses the information gained to harm his victim? Who will be held accountable then?

The NHTSA only says more research is needed on cybersecurity issues before finalizing regulatory standards. For now the agency is seeking to dodge and sidestep any debate over cyberspace. Instead, the NHTSA is calling on self-driving car manufacturers to directly share vulnerability data with their rivals through a medium known as Auto-ISAC, acting as a clearinghouse for the cybersecurity industry.

Government agencies are forbidden by Auto-ISAC to act privy regarding the disclosure of their vulnerabilities. Thus, the federal government can be set aside of the cybersecurity picture altogether. However, considering the growing worries over intelligence agencies hoarding zero-day vulnerabilities, along with the seemingly endless legal challenges regarding hacking by the government, keeping the government in the dark over any information on tech loopholes in self-driving cars that allow one to hack the vehicle is considered actually a very good idea.

Conclusion

The debate on how, when and with whom self-driving car data should be shared with will remain on the table for months to come. Some good news is that the NHTSA remains open to any feedback from industry stakeholders regarding its policy. However, actually convincing the entire industry and all their customers to reach an agreement with the Department of Transportation on how to approach the idea of data sharing is quite a challenge, to say the least.

Watch this video on how a world with self-driving cars would look like

Emerging questions: Are self-driving cars safe or at least safer than human driven cars? When will self-driving cars be available and affordable? Will self-driving replace completely traditional driving or  be limited to a drivers assistance role? Will the introduction of driverless car technology completely disrupt the automobile industry? Will the data collected and used for driverless cars end up being used for other purposes such as police surveillance or spying on people? Will hackers manage to get access to the information generated by autonomous cars? Could they take control of cars and disrupt traffic remotely? 

Vote to see result and collect 1 XP. Your vote is anonymous.
If you change your mind, you can change your vote simply by clicking on another option.