Abstract
In recent years, a shortage of commercial transport drivers has resulted in an inability to meet increasing demand. A proposed solution for combating this bottleneck is to automate long-haul trucks through sophisticated AI techniques. Through automation, this shortage can be resolved. Unfortunately, although there are too few truckers on the road, it remains one of the most common professions for Americans today. Thus, by embracing this technology, there is a risk of forcing a substantial portion of the population out of work. Would pursuing automated trucking be an ethical decision? Inversely, would it be unethical to disallow automation? Through various lenses, this paper will dive into this dilemma and determine the best path forward.
Introduction
In today’s era of technological advancement, the scribbles of science fiction become reality, and everyone waits to find out if authors like Philip K. Dick were justified in their paranoia. The public watches as more industries intersect with big tech, wondering if the embedded, artificially intelligent byproducts benefit the masses more than they cost them. The latest, and perhaps most controversial, occupation sitting idle at the big-tech intersection is long-haul trucking. The idea is that drivers can be replaced by upgrading to autonomous vehicles (AV) to increase productivity, safety, and sustainability. Truckers, along with many others, are not in support of this transition, and for good reason. Whereas autonomous freight (AF) will provide the public with a great deal of utility, this change disproportionately impacts the low and middle-class workforce. Therefore, while this inevitable transition shows significant promise, it also remains ethically unsound for engineers to develop AI trucks on their own. Instead, regulatory bodies should be established to consider the ethical implications of automation and to ensure that the burden as well as the benefits are more equally distributed.
Overview of the Trucking Industry
For the last few decades, the trucking industry has employed more Americans than almost any other. In 2021, estimates suggest that approximately 8.9 million people were employed within the trucking industry, including 3.5 million drivers on the road [1]. These Americans perform a critical role in the supply chain, and as such were listed as “essential workers” when the pandemic struck [2]. The profession is often described as grueling, under-salaried, and dangerous. The work week revolves around an 8-day period where operators can be expected to be behind the wheel for up to 70 hours [3]. Unfortunately, this does not factor in the time spent away from home waiting for loading, offloading, and maintenance. Though demand for drivers has steadily increased over the years, the earning potential has continued to decline. On average, career big-rig operators in the US bring home $60,161 before taxes annually, and the Department of Labor describes truck driving as one of the most dangerous occupations [4]. Coupling the non-standard hours with the difficulty of towing heavy loads, it is clear why accidents happen frequently and often result in death or injury. Additionally, because of the sedentary nature of the work and the inability to consume nutritious food on a regular basis, truckers incur numerous health issues [5].
Recently, the American Truckers Association (ATA) reported a shortage of as many as 80,000 long-haulers, and a majority of the workforce is in the process of aging into retirement. Today, the average US driver is a 46-year-old male. Although older generations endured uncomfortable living conditions, the compensation historically provided a greater incentive to work. Because wages have diminished with time, the industry is struggling to recruit and retain young workers [6]. Another issue is that trucking culture has always regarded itself as predominantly male and this stigma continues to act as a barrier to entry for women in industry.
Truck drivers serve a vital role in the supply chain, so it follows that their industry can significantly help or hinder economies. In recent years, supply has witnessed several bottlenecks that have affected everything from slowing aid relief during the pandemic to keeping toys off shelves leading up to Christmas. One explanation for these stocking deficiencies is that exponentially growing populations are creating an intractable demand. In addition, the gap between the number of available drivers and the number of drivers needed widens annually. As a result of these supply disruptions, the White House responded with a “trucking action plan” [7]. The plan involved a reduction of barriers to licensing, expanded opportunities for apprenticeships, and refocused outreach efforts to underrepresented demographics. Additionally, it called for a joint initiative between the Departments of Labor and Transportation to reevaluate the industry’s working conditions. Regrettably, forecasts suggest that this may still not be enough. Given this outlook, it is especially important to consider the numerous advantages artificial intelligence presents to long-haul trucking.
Benefits of AI – Utilitarian Case
Autonomous solutions offer many optimizations for the shipping industry without leaving blue-collar workers as far removed from the equation as many anticipate. Proceeding with the development of AF would stimulate the economy, increase public safety, and promote sustainability. From a purely utilitarian perspective, the decision to proceed is unambiguously clear.
Along their routes, truckers require stops for food, sleep, and restrooms that limit their productivity. By contrast, machines are not constrained by these physical limitations, nor federal regulations, so they can ship well beyond 70 hours each period. Though human involvement will still be necessary for performing roadside maintenance as well as urban driving, truckers will no longer endure the physical demands of being on the highway.
Another improvement AF promises to deliver is public safety. In 2019 alone, there were over 4,000 deaths and 130,000 injuries involving commercial big rigs [8]. When these accidents are fatal, it is often those riding in passenger vehicles who suffer the consequences [9]. From 2009 to 2017, these incidents increased by 52 percent [2]. By sidestepping the issues of fatigue and human error, AF will operate as vigilantly on its thirteenth hour of driving as its first. Moreover, it will proceed at reasonable speeds regardless of delivery pressures. Using a suite of cameras and radars for 360 degrees of “vision,” trucks will be able to hold steady within the confines of their lanes and maintain safer standoff distances from surrounding traffic [10].
In some ways, AI has already served the interests of trucking companies. Part of the appeal of artificial intelligence is that it can process large data sets and derive accurate predictions that would otherwise prove unsustainably tedious for humans. There is some proof of this in the routing performed by modern navigation apps such as Waze, Google, and Apple Maps. The route optimization employed by commercial shipping is even more powerful. These commercial algorithms also carry out the management and planning for entire fleets that save fuel and time by avoiding idle periods caused by congestion. A 2016 report estimated that it added 1.2 billion hours or 74.5 billion dollars to annual operations [11]. Further, by upgrading to more intelligent vehicles, a Berkeley study predicted as much as 35 billion in fuel efficiency gains [12]. Likewise, trucks with integrated AI can detect underlying mechanical issues that hurt performance [13]. In doing so, they can prevent down periods, reduce maintenance costs, and even save lives.
As a direct result of AI’s efficiency, the process of shipping can yield far greater productivity at a fraction of the current cost. This translates to public utility in better-stocked shelves, smaller price tags, and less carbon emissions. By coupling the faster processing power innate to computers with real-time sensory data, it will also be able to identify hazards and respond more quickly. Lastly, as intelligent agents, autonomous vehicles will evaluate sensor data to prevent incidents like tire blowouts. For these reasons and many more, AI trucks can force the upward-trending crash counts toward an inflection point in the years to come.
Consequences of AI – Fairness Rebuttal
While this transition in technology affords many opportunities for enriching our overall quality of life, it is not without its pitfalls. Two reasonable concerns are that automation will eliminate jobs and that the bias programmed into intelligent agents cannot be implemented in a manner that is unanimously agreeable. From this perspective, the decision to automate trucks is unfair and thereby unethical.
Trucking represents one of the largest means of employment for Americans. When the day comes that automated freight has been adequately tested, companies will begin to buy in. Millions may go unemployed in the US and such a crisis could create more economic harm than good. Even with established retraining programs, drivers may still find it difficult to flex into new roles due to their average age and education level [6].
This is not the first time an industry’s workforce has been negatively impacted by technology. Just a few decades ago, the primary employer for blue-collar workers was manufacturing [14]. Following the incorporation of automation through robotics in factories, many workers were pushed out of the industry into fields such as trucking. In 2016, a management and consulting company, McKinsey, sought to gauge how automation through robotics affected the US Labor market. They focused specifically on shared markets between neighboring rural and urban communities, known as commuting zones, and utilized data from the International Federation of Robotics (IFR). Ultimately, they estimated that the addition of one robot reduced annual wages by $200 and replaced 6.2 workers in each community zone [15]. Meanwhile, companies that invested in robots saw substantial gains in revenue and productivity. Hence, the driverless movement has the potential to disproportionately burden the lower classes and breed greater economic inequality in society.
Another angle that remains widely controversial is bias in AI platforms. Regarding autonomous vehicles, the most discussed topic surrounds what is known as the “Trolley Dilemma,” first posed by philosopher Philippa Foot [16]. In this circumstance, an individual (or intelligent entity) must decide between the lives of some over others. Though engineers hope to dramatically reduce death and injury counts linked to long-haul trucks, they cannot altogether avoid situations of imminent collision. Therefore, they must train autonomous agents to take in their predicament, evaluate for possible outcomes, and react with bias. For example, should an AV aim to save the driver or those it is colliding with? One person over multiple, felon over retiree, or one race over another? There seems to be no clear-cut way to proceed with one framework. Moreover, it is utterly incalculable to consider each instance in a matter of milliseconds.
Ultimately, from the perspective of fairness, the widespread incorporation of autonomous shipping is ethically unsound. The decrease in price tags cannot offset the financial loss those without work would experience. No matter the bias adopted for extreme situations and edge cases, it will likely still be unfair to someone.
Rights-based Tiebreaker
So far, the pros and cons corresponding to driverless trucks have shown polarized responses from differing ethical standpoints. On one end, utilitarian analysis suggests that the benefits to people as a collective should incentivize such a transition. At the other end of the spectrum, fairness-based ethics highlights the harm AF can introduce. Perhaps the most sensible response should be to entertain more frameworks and observe how the scales shift.
Due to the many enhancements automation provides, it may also be ethical from a rights-based approach. Given the disruptions in the current supply and worrisome forecasts of future breakdowns, the public has a democratic right to turn to AI solutions. Additionally, in our capitalistic system, one could claim that big tech has a right to innovate pragmatic solutions. A core of the American dream is the freedom to pursue success at the risk of failure. However, the question must be posed: at which point does success become excess?
Another perspective suggests that automating trucking will violate some individuals’ rights. Although engineers have the right to innovate, that right should not conflict with or infringe upon the rights of others. Rights such as the ability to work and provide decent living circumstances need also be factored into the development of this technology. Surely, entire workforces should not be left suddenly unable to pay rent or purchase groceries. Additionally, truckers’ suffering is not limited solely to their wallets, but also to their mental health. Losing one’s ability to provide and a reason to get up in the morning can be staggering.
Viewing the problem from a middle ground offers insight that can help smooth an otherwise difficult transition. This viewpoint also cements the idea that engineers alone cannot correct the deficiencies in trucking. If big tech has the right to innovate for society’s betterment, it might also be said that they must catch those they knock down.
Conclusion
Artificial intelligence appears to be the answer to many modern problems. The field is growing rapidly because technology is polymorphic and yields astonishing results, such as the improvements it can make for the trucking industry and the supply chain. Nevertheless, viewing AI’s further integration into long-haul shipping from varying ethical perspectives illustrates how not everyone benefits from it, including the many people who will be left without income and opportunity. Although engineers often have good intentions, they cannot reasonably be left to solve such pervasive problems without diligent consultation of everyone involved. Nor can artificial intelligence be continually pushed into public service without a thorough analysis of its implications.
By Jacob Harrington, Viterbi School of Engineering, University of Southern California
About the Author
At the time of writing this article, Jacob Harrington was a senior studying Computer Science at the University of Southern California. While fascinated by both the opportunities that Artificial Intelligence affords society and the process involved in innovating such intelligent agents, he remains skeptical of the side effects it unintentionally generates.
References
[1] D. Cullen, “ATA: Truck driver shortage hits record high of 80,000,” Drivers – Trucking Info, 26-Oct-2021. [Online]. Available: https://www.truckinginfo.com/10154800/ata-truck-driver-shortage-hits-record-high-of-80-000#:~:text=The%20American%20Trucking%20Associations%20estimates,in%20less%20than%20ten%20years (accessed Mar. 31, 2022).
[2] “Technology,” Embark Trucks, 13-Sep-2021. [Online]. Available: https://embarktrucks.com/technology/ (accessed Apr. 7, 2022).
[3] “How many hours can a truck driver drive?,” Samsara, 03-Nov-2021. [Online]. Available: https://www.samsara.com/guides/how-many-hours-can-a-truck-driver-drive/ (accessed Apr. 7, 2022).
[4] “Truck+driver salary in USA – average salary,” Talent.com. [Online]. Available: https://www.talent.com/salary?job=truck%2Bdriver (accessed Apr. 7, 2022).
[5] J. Hirsch, “Federal data shows trucking remains one of America’s deadliest jobs,” Trucks.com, 13-Jan-2021. [Online]. Available: https://www.trucks.com/2021/01/11/federal-data-shows-trucking-remains-one-of-americas-deadliest-jobs/ (accessed Apr. 7, 2022).
[6] J. Kirby, “Where have all the truck drivers gone?”, Vox, 02-Jan-2022. [Online]. Available: https://www.vox.com/22841783/truck-drivers-shortage-supply-chain-pandemic (accessed Apr. 7, 2022).
[7] “Fact sheet: The Biden-Harris Administration Trucking Action Plan to Strengthen America’s Trucking Workforce,” The White House, 16-Dec-2021. [Online]. Available: https://www.whitehouse.gov/briefing-room/statements-releases/2021/12/16/fact-sheet-the-biden-%E2%81%A0harris-administration-trucking-action-plan-to-strengthen-americas-trucking-workforce/ (accessed Apr. 7, 2022).
[8] “Truck accident statistics 2021: Munley Law: Truck accident lawyers,” Munley Law, 21-Sep-2021. [Online]. Available: https://munley.com/truck-accident-statistics-2021/ (accessed Apr. 7, 2022).
[9] “Truck accident statistics 2021: Munley Law: Truck accident lawyers,” Munley Law, 21-Sep-2021. [Online]. Available: https://munley.com/truck-accident-statistics-2021/ (accessed Apr. 7, 2022).
[10] “Autonomous Truck Driving Technology,” Plus. [Online]. Available: https://plus.ai/technology.html (accessed Apr. 7, 2022).
[11] S. Ashcroft, “Top 10: Worst Supply Chain Disasters in history,” Supply Chain Digital, 07-Dec-2021. [Online]. Available: https://supplychaindigital.com/supply-chain-risk-management/top-10-worst-supply-chain-disasters-history (accessed Mar. 31, 2022).
[12] S. Viscelli, “Driverless? autonomous trucks and the future of the American trucker,” UC Berkeley Labor Center, 04-Aug-2021. [Online]. Available: https://laborcenter.berkeley.edu/driverless/. [Accessed: 25-Apr-2022].
[13] R. Dawson, “Implications of AI for the Trucking Industry,” Ross Dawson, 18-Jun-2020. [Online]. Available: https://rossdawson.com/futurist/implications-of-ai/implications-of-ai-for-the-trucking-industry/#:~:text=Fuel%20Efficiency%20and%20Savings&text=AI%20can%20be%20used%20to,will%20be%20cut%20by%2015%25 (accessed Apr. 7, 2022).
[14] Q. Bui, “Map: The most common* job in every state,” NPR, 05-Feb-2015. [Online]. Available: https://www.npr.org/sections/money/2015/02/05/382664837/map-the-most-common-job-in-every-state (accessed Mar. 31, 2022).
[15] I. Acevedo, Ivonne AcevedoIvonne (’17) is a staff writer for Labor & Finance. She is interested in public finance. More », S. Edwards, and A. Swimmer, “Do industrial robots contribute to unemployment and lower wages?,” Chicago Policy Review, 08-Jun-2017. [Online]. Available: https://chicagopolicyreview.org/2017/06/07/do-industrial-robots-contribute-to-unemployment-and-lower-wages/ (accessed May 9, 2022).
[16] W. Grimes, “Philippa Foot, renowned philosopher, dies at 90,” The New York Times, 09-Oct-2010. [Online]. Available: https://www.nytimes.com/2010/10/10/us/10foot.html#:~:text=Philippa%20Foot%2C%20a%20philosopher%20who,3%2C%20her%2090th%20birthday (accessed Apr. 7, 2022).