Research Predicts Adoption Rates of Connected Autonomous Vehicles
In this installment of the Innovation of the Month series, we explore how the FedEx Institute of Technology at the University of Memphis, in partnership with the city of Memphis, has developed a methodology for predicting connected autonomous vehicle adoption.
MetroLab’s Executive Director Ben Levine discussed the project with Sabya Mishra, co-director of the SMARTCities Research Cluster at the University of Memphis, and Memphis CIO Mike Rodriguez.
Ben Levine: Your project focuses on modeling people’s adoption of autonomous vehicles, which seems to mix the practical reality that autonomous vehicles are coming with the complex issue of how potential users will actually react to and use the technology. Can you talk about that dynamic?
Sabya Mishra: You are right that connected autonomous vehicles (CAVs) are about to become a reality, and they are arriving much earlier than many would think. By incorporating features such as parking assist, adaptive cruise control and collision avoidance systems, most automobile manufacturers have already incorporated some degrees of automation into the existing cars. Mercedes-Benz, Google, Tesla and others have already developed and tested prototypes of the first fully autonomous vehicles. By aggressive testing of autonomous peer-to-peer ridesharing services, transportation network companies such as Uber and Lyft are also pushing the introduction of automation.
One key question that has been of interest to policymakers, academic researchers and industry professionals is how much demand there will be for ownership of CAVs and what will be the timing of adoption in the long term? Our project attempted to address this question by collecting survey data, and applying Diffusion of Innovation theory and with agent-based modeling to forecast CAV adoption.
Mike Rodriguez: The city of Memphis remains focused on decreasing traffic congestion and encouraging public transportation by providing affordable options. CAVs can support these efforts if we are proactive in developing a policy framework that provides guidance and reduces potential barriers. This project serves to provide greater insight into the individual demand and the forecasted adoption timeline. With these insights we can better prepare for infrastructure modifications and/or upgrades. Additionally, it highlights an opportunity to create a network to provide low-income citizens transportation services to and from medical visits, which frees up our emergency vehicles.
Levine: Can you share more about the idea of Diffusion of Innovation?
Mishra: The theory of Diffusion of Innovation seeks to understand how an innovation will diffuse as a result of communication and consumer interactions in a social network. It considers innovation diffusion as a social phenomenon that has four aspects: 1) the demand to adopt the innovation; 2) communication through certain channels; 3) communication among individuals in a social network; and 4) communication over time.
Diffusion research has been of interest to the academic community for an extended period of time. The seminal diffusion study was published in 1943, when Bryce Ryan and Neal Gross modeled the diffusion of hybrid seed corn among Iowa farmers.
Everett Rogers, a professor of communication studies, and others then extended the DOI theory application in many disciplines such as health care, computer science, engineering technology, agriculture, information technology, electrical and electronics engineering, etc. Our study extends its application to study the adoption of autonomous vehicles.
Levine: What do you think is embedded in someone’s consideration of adoption of connected and autonomous vehicles?
Mishra: Well, this is really what our analysis looked at. To give some examples, we conducted a survey that resulted in some interesting findings. For example, less than 5 percent of respondents state that their households are willing to pay an additional $20,000 to add automation and connectivity; 69.1 percent of respondents' households are willing to pay only an additional $5,000 or less to have the driverless option added to their car. We found that improving social status among peers is the least important incentive to adopting CAVs, and the risk of virus attack is considered the most important barrier to adoption. One very interesting finding was that the losing feeling of control was more important than safety concerns for people.
Figure 1: The graph above shows the Resistance to Self-Driving Car Adoption. Courtesy of Sabya Mishra.
Then, embedded in our analysis — and DOI modeling generally — is the notion that a person’s choices will have an impact on the choices of others. This conceptually makes sense — think about how your familiarity or comfort with a technology is shaped by those around you. In particular, we look at how positive and negative perceptions about CAVs will actually have an impact on somebody else’s willingness to pay for the technology. This modeling is undergirded by the concept of homophily, which indicates that individuals with geographical proximity and socioeconomic similarity are more likely to impact each other.
Levine: How will the findings developed in this project impact future planning?
Rodriguez: The data collected in this project supports initiatives such as smart parking, street lights and traffic sensors optimized for upcoming technology. Additionally, it will help inform city policies, budget, and planning efforts as it relates to infrastructure, particularly fiber deployment. As we continue to build out a Smart City strategy, the project serves as a template for which other research can be applied.
Mishra: FedEx Institute at University of Memphis is supportive of innovative and impactful research such as adoption of autonomous vehicles. In addition, now we have partnered with the TennSMART consortium, a collaboration between universities, cities and industries in Tennessee to explore more on smart and connected cities research. FedEx Institute is also a partner with the Association for Unmanned Vehicle Systems International to explore further nationwide opportunities. FedEx Institute also has a strong collaboration with MetroLab Network to further strengthen city-university relationships.
Levine: I have a question related to future transportation planning. It sounds like your analysis considered a CAV future in which people own their own vehicles. I think there is a broader question about how the vehicle market will look generally in a world with CAVs. We are already seeing that with increases in ride-sharing and car-sharing.
Mishra: This is a good question, and certainly we have limited information as of yet on how vehicle ownership will change over time. Especially, if the capital, operation and maintenance costs of CAVs are very high, then we may see people using more of ride-sharing and car-sharing. Also there is a potential for individuals leasing CAVs for their use; and paying the leasing amount per month rather than worrying about purchasing and maintaining the CAV. More research is needed in the domain of vehicle ownership versus ride-sharing to come up with certain plausible scenarios.
Levine: Mike, how are you thinking about this from a policy perspective? Do you imagine a future in which Memphians own their own CAVs? Or do you think you’ll see an increase in pooling, sharing, and fleets owned by the transportation network companies or vehicle manufacturers?
Rodriguez: If you consider the shift in business over the last decade, you see an obvious trend toward a shared economy. This sharing is made possible through the advancement of technology. In the last few years we’ve seen an increase in ride-sharing and pooling. If vehicle technology grows as it has in the past decade, we may see a trend toward a shared economy in this industry as well. Owning versus ride-sharing is an evolving question, one that we are continuing to research. In the next few years we will gain much-needed insight into the future of vehicle ownership.
Levine: Where will this project go from here?
Mishra: The project was instrumental in developing a foundation for autonomous vehicle adoption research. Three other projects are currently underway. One of the projects is exploring the development of a readiness index for U.S. smart cities, and the other two projects are exploring organizational adoption — such as by freight companies, and transit agencies — of autonomous vehicles. In addition, this study has provided impetus to extend the adoption research to the whole state of Tennessee, while a number of researchers and practitioners around the world are using foundational concepts developed in this paper to perform case studies in their respective regions.
Rodriguez: The research conducted in this project provides a solid foundation for continued exploration of mobility technology and its impact in city planning. We look forward to working with the University of Memphis to investigate other technological areas of innovation, especially as it relates to our smart city strategy.
Levine: How did this project fit into the broader strategy on smart cities and urban innovation at the University of Memphis?
Mishra: FedEx Institute of Technology at the University of Memphis developed a research cluster on “SMARTCities.” As a part of MetroLab Network’s university and city partnerships, the SMARTCities research cluster focuses on emerging and innovative research topics that have potential for innovation and implementation to help smart city efforts at the city and national level. A set of diverse topics from six different disciplines were funded to conduct research in the broader aspects of smart cities.