Autonomous bicycles have recently been proposed as a new and more efficient approach to bicycle-sharing systems, but how do they compare from an environmental perspective?
Autonomous bicycles have recently been proposed as a new and more efficient approach to bicycle-sharing systems, but how do they compare from an environmental perspective?
At the City Science group, we have developed autonomous shared micro-mobility systems, such as the MIT Autonomous Bicycle Project, intending to bring mobility-on-demand convenience into already prevalent bicycle-sharing systems. Increased convenience could incentivize more people to ride, reducing car dependency and making cities more livable and human-centric.
Autonomy could help alleviate some of the challenges faced with current bike-sharing, such as the rebalancing problem or the bicycle oversupply. Since autonomous bicycles would be able to relocate by themselves, system operators would no longer need to redistribute bikes from bicycle-saturated areas in rebalancing vans. Similarly, the user experience could also be improved by providing increased predictability and reliability and eliminating the need to find available bicycles or docks. In addition, being an inherently more efficient system, the number of bikes needed for a certain demand would also be smaller, reducing the problem of bicycle oversupply.
A previous simulation study shows that, with a fleet size 3.5 times smaller than a station-based system and eight times smaller than a dockless system, an autonomous system can improve overall performance and user experience.
Knowing that autonomy could improve the efficiency of current bicycle-sharing systems, we also wanted to compare them from an environmental perspective. Conducting environmental impact studies at an early technology development stage is critical because it is when there is the greatest opportunity to influence the design and, ultimately, environmental impacts of a system. However, this stage is challenging due to the lack of real-world data and associated uncertainty. While our study tries to limit these uncertainties by focusing on broad scenario studies and sensitivity analyses, it should be kept in mind that environmental impacts are very case-specific.
For the nominal scenario considered in our study, the CO2 equivalent emissions per kilometer traveled with autonomous shared bicycles would be 33.1 % lower than in station-based systems and 58.0 % lower than in dockless systems. This reduction in environmental impacts is mainly driven by the smaller needed fleet sizes and the elimination of bicycle rebalancing vans.
The sensitivity analysis shows that the environmental impact of autonomous shared bicycles will mainly depend on the levels of use and infrastructure. In addition, other design variables such as increasing vehicle lifetime and reducing the components needed for autonomous driving will also significantly reduce the environmental impacts of autonomous shared bicycles. The improvement in the environmental performance of autonomous shared bikes over current systems becomes especially significant when accounting for the displaced mobility modes. Currently, most bike-share trips substitute walking, private bikes, or new trips induced by the BSS. Therefore, it will be critical to consider that, as other mobility modes decarbonize and become more efficient, shared bicycle systems should focus on targeting mode replacement from more polluting modes.
We hope that by identifying the parameters that have the most significant influence on the environmental impacts of autonomous shared bikes, this study will assist vehicle design engineers, system operators, urban planners, and governments in their respective decision-making processes. As these systems are further developed, we expect the insights from this study to help prioritize the data gathering of the most critical variables from a sustainability perspective. As more data becomes available on how these systems will behave, reducing the uncertainties described in this study, their environmental impacts should be reevaluated.