Disadvantages of Autonomous Vehicles

Autonomous vehicles, also known as driverless cars, are cars with the capability of sensing their surrounding environment and making decisions, allowing for navigation without human input. There are many different conversations and opinions happening today in the world of driverless cars. These vehicles are becoming increasingly widespread in our society due to their features that theoretically make driving safer, easier, and more efficient. Over 90% of accidents are caused by human error (United States Congress) and eliminating that variable could potentially save thousands of lives. In addition, these cars can make driving easier by increasing the value in travel time and decreasing fuel use (Van Den Berg). Even though there are many great positives to these vehicles, there are impediments blocking their growth. It is difficult for the mass public to get behind these cars, because no company can clearly address many of the underly issues that these cars bring forward.

Many vehicles being released today already have autonomous features, without being fully automated. Autonomous features are the attributes of cars that allow for a vehicle to make decisions on its own. These features do not require any human action to be taken for a specific task. However, the vehicle still is not fully autonomous and must have a human in control much of the time spent driving. Vehicles can be classified by their respective number of autonomous features. The more autonomous features the higher the level vehicle falls under (Table 1). Level zero would be no automation and level five would be a fully automated vehicle. Anything under a level three would require mostly human input with only help from things like power steering and lane change assist. A level three and above is almost always controlled autonomously with the differences being in whether the car needs human input in the case of an accident (Borenstein).

Table 1

Autonomous vehicles have massive implications making the issue of implementing them in our world both puzzling and problematic. The problems that cars bring up can be broken down to ethical, legal, or economical dilemmas. The potential detrimental effects of these cars do not typically fit into one specific category, because the issues are complicated and overlapping. Instead of focusing my research on only one area of the issue, I have decided to take a broader approach and look at the many implications of implementing these cars. Today we are still uncertain of appropriate solutions to the difficulties brought about by autonomous vehicles. As a variety of companies continue to see the great positives that come from autonomous vehicles, they will continue to push to get these cars on the road. We must be ready to face the problems of these vehicles with potential solutions.

 It is tough to predict the exact implications of autonomous vehicles until they are fully incorporated into daily life. It is difficult to predict if autonomous vehicles will live up to all the claims being made by the manufacturing companies. It is quite likely that new complications will arise as time goes on. While autonomous vehicles may eliminate an issue such as human error, we made find that these vehicles end up glitching and making more mistakes than an average human. In the current age, there are a few main areas of concern. These main concerns being raised involve ethics, safety, and economics. There are the ethical concerns involving accident algorithms and responsibility of an accident. There are safety concerns of these vehicles being hacked into and hijacked. Lastly, there are economic concerns of loss driving related jobs, such as a bus driver or uber driver, and the effect these vehicles will have on public transportation.

An accident algorithm is data programmed into an autonomous vehicle, deciding how the vehicle responds during an unavoidable accident (Nyholm). The issue with this type of data is that a programmer would be making tough decisions as to who should be injured in an accident. It gives the programmer a lot of responsibility and makes the programmer put a value on each of the passengers in a car. A common scenario that many ethicists use when talking about driverless vehicles is the “Trolley Problem.” This experiment asks people what they would do if a trolley is barreling down a track toward five people, and you have the opportunity to save those five lives by changing the course the trolley to where only one person stands (Thomson). Ethicists see a parallel between this and the accident algorithms of autonomous vehicles “…should autonomous vehicles be programmed to always minimize the number of deaths? Or should they perhaps be programmed to save their passengers at all costs?” (Nyholm). Programmers would be making a choice between the lives of the passengers of the driverless cars and potentially saving the lives of many civilians. The clear distinction between what happens today in accidents and accident algorithms is that today when an accident occurs we consider the reaction of the driver to be just that, a reaction. An accident algorithm is pre-programmed to decide who dies and therefore it is much closer to premediated murder. Yet, this complicated issue of how much power we are giving to programmers can be solved using laws. Washington D.C. has passed legislation that requires a human driver to be prepared to take control of the autonomous vehicle at any moment (National Congress). Additionally, in Germany, a similar law has been passed where “the driver will bear responsibility for accidents that take place under his or her watch…but if the self-driving system is in charge and a system failure is to blame, the manufacturer will be responsible” (Wacket). This means that the issue of the trolley problem and power given to programmers is no longer important. The car would not make the decisions in an unavoidable accident, it would be the driver who takes over. If this law was enacted nationally it would allow for programmers to create these cars without the fear of how they should program the cars in unavoidable accidents.

Moreover, an issue with Autonomous vehicles that these cars will make it impossible to find fault for an accident. The idea that a machine cannot be held responsible is not incorrect, but it isn’t the full story. “There can be always either a human being or a legal entity held responsible and liable for the actions of the AICs. (De Cock Buning)” This means that blame may not be able to be placed on the robot itself, but similar to other AI technology the responsibility of the actions of the AI, falls on both the coder and designer. The designer being the person that comes up with the idea and plan for what the AI should do and how many features should be autonomous. The designer decides how the AI should act in different situations. The coder executes the demands of the designer and creates the framework for the decision making of the cars. The coder and the designers can work in teams but no matter what the people that create and develop the vehicles will always be at fault for how they act. Therefore, there should not be an issue on fault for these accidents. The designers and coders are just going to have to accept fault and risk for their creations because the AI itself cannot be blamed.

Another common problem that people find with autonomous vehicles is the issue of hacking and security of these cars. There are people that believe that these cars lack in security because they may be used for terrorist attacks and they may be easily hacked into (Litman). While this issue of security is difficult and complicated, there may be a way to solve these issues with laws. One example of this is with mandatory penetration testing before a car is released. Penetration testing is controlled hacking on a system to find its weaknesses (Arkin). The importance of this testing is that it finds weaknesses in the software before it is released to the public, so that programmers have time to fix the issues. This will create better, more secure software that is harder to hack. In addition, if there were policies that incentivized the sharing of information between companies, then the cars could be safer overall. These companies are naturally competitive, but if customers do not feel safe in any autonomous vehicle, then they are going to have to work together to improve their systems.

Yet another problem people have with autonomous vehicles is that they may distract attention and consequently lower funding for mass transit which would hurt the poor (Fleetwood). Yet, the issue with this concern is that it is putting the blame on the vehicles rather than the people in government. Policymakers will be held accountable for their decisions on where to spend the United States budget. Today, there are already programs that help public transportation to be available to take people to their jobs. An example of such a program is the Job Access and Reverse Commute Project (JARC program). This is open to eligible low-income individuals who need transport to and from jobs. This project “facilitates the provision of public transportation services from urbanized areas and rural areas to suburban employment locations” (APTA). This means that it would be extremely difficult to significantly cut down on public transportation, when there is a law requiring that people have access to their jobs. Moreover, driverless cars will decrease the cost of ride sharing services because there is no longer a need to pay a driver. These cars should also decrease the cost of riding on public transportation considering you may not have to pay for a driver on your city bus. Therefore, public transportation will exist, but may exist in a new fashion.

One consistent concern with driverless cars is the issue of lost jobs. People who have jobs that revolve around driving, such as taxi or Uber drivers, will be in jeopardy of replacement by autonomous vehicles “between now and 2030, driverless features will pave the way for fully autonomous vehicles and the demand for drivers will begin to plummet” (Fray). As these cars become integrated into our daily lives, jobs will be lost. However, this is not true just for the automation industry. In multiple industries there is worry of a phenomenon called technological unemployment. This happens when an increase in technology causes for the unemployment of workers. This relationship between technological progress and unemployment has been a concern of both economic theorists and policy makers, since the beginning of industrialization in the late 18th century (Postel-Vinay). Two economists from MIT, Erik Brynjolfsson and Andrew P. McAfee, claim that automation is replacing people faster than the economy can create jobs. Brynjolfsson and McAfee agree that to succeed economically we must work with computers rather than against them. “Computers are better at repetitive, error-free routines, and are improving in pattern recognition and complex communication. Humans have the intuition and creativity computers lack. They can define their own tasks and work outside their predefined domains. Married by the right processes, people and machines could forge potentially beautiful partnerships” (Brown). A law that would help people to learn work with technology would be a type of retraining program for dislocated workers. There was a similar law to this 2014 that would help dislocated workers achieve job retraining and a valuable skillset. After an analysis of the Workforce Innovation and Opportunity Act (WIOA) effectiveness, it has been determined that WIOA does help “unemployed to upgrade their skills and improve their prospects in the workforce” (Pedro). I would suggest that we work on this policy to directly apply it to drivers who will lose work because of these new cars. This way we use the effective resources and programs that we already have to retrain these employees for better jobs.

Autonomous vehicles effect all our lives. Whether or not you plan on buying a vehicle they are expected to be on the road in a very short amount of time. The way these cars handle themselves in accidents and on the roads put all drivers at risk. We need to be aware of all of the risks of the cars and have ideas on how to fix some of the problems. We cannot be passive about the arrival of autonomous vehicles. We must be prepared for these cars and make legislation ahead of time to counteract some of the potential problems of these cars. Other companies are already working on advanced stages of this technology and if we are too afraid to make the proper legislation, then we will not be able to have these cars on the roads and America will fall behind.

A common issue with these cars is if they are pre-programmed with how to deal with a crisis, then some people may consider the outcome of the car’s decision to be premeditated murder. In addition, while these cars can make us more productive and save lives, they may be at risk for hacking, so they may not be necessarily safer.  In addition, there are economic concerns such as how these cars will affect public transportation for poor America and jobs that will be lost to these cars. While companies such as Google, Tesla, and BMW push to create these vehicles, many economists and philosophers have become increasingly concerned about the power we are giving to autonomous vehicles. Overall many problems of autonomous vehicles can be solved by laws. This will hopefully allow us to enjoy the benefits of driverless cars without the worry of some of the major concerns. It is imperative that we work on legislation before these cars are allowed on the roads because there are human lives at risk. If we institute the laws such as mandatory penetration testing, drivers’ ability to take over, and affordable public transportation, then we can create a less controversial environment that these cars could be instituted and exist in.

Works Cited

• APTA. FEDERAL PUBLIC TRANSPORTATION LAW. American Public Transportation Association, www.apta.com/gap/legissues/authorization/Documents/Ramseyer_Ch_53_Revisions_Final%20(2).pdf.
• Arkin, Brad, et al. “Software Penetration Testing.” IEEE Security & Privacy , vol. 3, no. 1, 14 Feb. 2005, pp. 84–87.
• Borenstein, Jason, Joseph Herkert, and Keith Miller. “Self-Driving Cars: Ethical    Responsibilities of Design Engineers.” IEEE Technology & Society Magazine, vol. 36, no. 2, 2017, pp. 67-75.
• Brown, Alan S. “Automation Vs. Jobs. (Cover Story).” Mechanical Engineering, vol. 134, no. 4, Apr. 2012, pp. 22-27. EBSCOhost, mutex.gmu.edu/login?url=http://search.ebscohost.com/login.aspx?direct=true&db=mth&AN=73911437&site=ehost-live.
• De Cock Buning, Madeleine, and De Bruin, Roeland. “Autonomous Intelligent Cars: Proof that the EPSRC Principles are Future-Proof.” Connection Science, vol. 29, no. 3, 2017, pp. 189-199, ProQuest Central; ProQuest SciTech Collection; Research Library,
• Fleetwood, Janet. “Public Health, Ethics, and Autonomous Vehicles.” American Journal of Public Health, vol. 107, no. 4, 2017, pp. 532–537.
• Fray, Thomas. “101 Endangered Jobs by 2030.” Ebscohost, Journal of Environmental Health, 1 May 2015, web.a.ebscohost.com/ehost/pdfviewer/pdfviewer?vid=1&sid=2c7fb6d6-1e17-4e6d-923e-9b891cb0640f%40sessionmgr4010.
• Litman, Todd. “Autonomous Vehicle Implementation Predictions.” Victoria Transport Policy Institute, 27 Feb. 2017.
• López-Claros, Augusto. The Innovation for Development Report 2009-2010: Strengthening Innovation for the Prosperity of Nations. Palgrave Macmillan, 2010.
• National Congress of State Legislatures. Autonomous Vehicle Enacted Legislation. 16 Oct. 2017, www.ncsl.org/research/transportation/autonomous-vehicles-self-driving-vehicles-enacted-legislation.aspx.
• Nyholm, Sven, and Jilles, Smids. “The Ethics of Accident-Algorithms for Self-Driving Cars.” Ethical Theory and Moral Practice, vol. 19, no. 5, 2016, pp. 1275–1289.
• Pedro, Chuck. Job Training Programs for Dislocated Workers. John Hopkins University, Dec. 2016, jscholarship.library.jhu.edu/bitstream/handle/1774.2/40169/Chuck%20Pedro.pdf? sequence=1&isAllowed=y.
• Postel‐Vinay, Fabien. “The Dynamics of Technological Unemployment.” International Economic Review, Blackwell Publishing, 6 Aug. 2002, onlinelibrary.wiley.com/doi/10.1111/1468-2354.t01-1-00033/full.
• Rotman, David. “What Role Does Technology Play in Record Levels of Income Inequality?” MIT Technology Review, MIT Technology Review, 1 Sept. 2016, www.technologyreview.com/s/531726/technology-and-inequality/.
• Thomson, Judith. “The Trolley Problem.” The Yale Law Journal, vol. 94, no. 6, May 1985, pp. 1395–1415.
• United States Congress, U.S. Department of Transportation, National Highway Traffic Safety Administration. “National Motor Vehicle Crash Causation Survey: Report to Congress.” National Motor Vehicle Crash Causation Survey: Report to Congress, U.S. Department of Transportation, 2008, pp. 1–47. crashstats.nhtsa.dot.gov/Api/Public/View Publication/811059.
• Van Den Berg, Vincent A.c., and Verhoef, Erik. “Autonomous Cars and Dynamic Bottleneck      Congestion: The Effects on Capacity, Value of Time and Preference Heterogeneity.” Transportation Research Part B: Methodological, vol. 94, Dec. 2016, pp. 43–60.
• Wacket, Markus, and Thomas Escritt; “Germany Adopts Self-Driving Vehicles Law.” Reuters, Thomson Reuters, 12 May 2017, www.reuters.com/article/us-germany-autos-self-driving/germany-adopts-self-driving-vehicles-law-idUSKBN1881HY.