Poor travel-time reliability (TTR) in our cities means that travel times for the same journey are both highly variable and unpredictable, and also gives rise to similar negative impacts on the environment and the economy as does traffic congestion. Being able to offer a high degree of TTR is therefore highly desirable for travellers and will also facilitate the uptake of sustainable urban transportation including future public, shared, and on-demand mobility services, and on-time delivery of freight.
Connected and automated vehicles (CAVs), such as so-called robotaxis and automated shuttles, are now beginning to be deployed in our cities [1] and their widespread deployment might help to improve TTR. Currently, however, they operate largely independently as replacements for conventional human-driven vehicles (HDVs) and, indeed, their presence may even have a negative impact on traffic. This project will therefore explore how collections of CAVs might be orchestrated to harmonize traffic flow (including HDVs) and improve TTR. The project is expected to build on our existing work on the use of reinforcement learning for optimizing TTR in the absence of HDVs.
[1] https://www.connectedautomateddriving.eu/observatory/observatory-tables/services/, accessed 2026-08-19
[2] Mass robotaxi malfunction halts traffic in Chinese city, accessed 2026-08-19