Optimizing Urban Transit through Passenger Spatial and Temporal Flexibility

This direction is merged into another direction: spatial demand imbalance. Its last version stays readable and citable here.

Version 1 of 114 papers22 conceptsfrozen 2026-10-04

Traditional transit systems treat passengers as "passive cargo"—static points on a map that must be collected and delivered to exact coordinates. This "door-to-door" rigidity creates a massive computational and operational burden, forcing vehicles into long detours and "empty cruising" miles.

However, a growing body of research suggests that the key to unlocking urban mobility efficiency lies in treating passengers as active participants. By introducing small amounts of flexibility—such as walking a short distance to a "virtual stop" or accepting an alternative destination—the system-wide benefits scale disproportionately to the individual effort required.

The Geometry of the Walking Buffer

The most fundamental mechanism for improving efficiency is the introduction of a small walking buffer. In traditional ride-sharing, a vehicle must deviate from its optimal path to reach a passenger's exact doorstep. As Hao Jiang and Pradeep Varakantham demonstrate in their work on FlexiPool, allowing passengers to walk to optimized meeting points can slash average travel distances by 21%.

The intuition here is geometric: a single point is a rigid constraint, but a 100-meter radius around that point is a region of possibilities. When an algorithm can pick any point within that circle to intersect a vehicle's path, the likelihood of finding a "zero-detour" pickup increases dramatically. Dániel Kondor and Carlo Ratti found that in autonomous micromobility networks, a willingness to walk just 100 meters can reduce the necessary fleet size by approximately 25%.

From Fixed Points to Virtual Hubs

To manage this flexibility at scale, researchers have moved toward "virtual stops"—dynamically assigned locations that balance user convenience with fleet throughput.

In the MARTA Reach pilot, Pascal Van Hentenryck and colleagues showed that these hubs allow on-demand shuttles to function as efficient "feeders" for fixed-route rail lines. Similarly, Jiachuan Wang and Lei Chen proposed the Online Ridesharing with Meeting Points (MORP) framework. By using a graph-theoretic structure called a $k$-skip cover, they can quickly identify meeting points that serve the maximum number of riders with minimum vehicle deviation, serving up to 33% more requests than door-to-door models.

Interestingly, the flexibility of the drop-off point often yields higher gains than the pickup point. As noted in the FlexiPool study, the drop-off location determines where the vehicle will be for its next request. Optimizing this "end state" creates a compounding effect on fleet distribution across the city.

Passengers as Proactive Agents

Beyond simple walking, researchers are exploring "prosocial" flexibility—where riders adjust their behavior for the benefit of the community. Baudouin Dafflon and Maxime Guériau describe this as the emergence of micro-communities, where riders self-organize to meet at shared locations, boosting occupancy and reducing wait times.

This shift from passive to proactive is even being applied to vertical transit. Hangli Ge and Noboru Koshizuka developed Intellevator, a system that negotiates with elevator passengers. By suggesting a user walk one floor up or down to skip a stop, the system reduces building-wide congestion.

In more complex social environments, Eleni Bardaka and Munindar P. Singh propose CARS (Cooperative Adaptive Ride-Sharing). Instead of using financial penalties like surge pricing, CARS uses AI-driven empathy-building messages to nudge flexible riders to change their schedules or walk further. This frees up capacity for "inflexible" riders—those with medical appointments or strict work shifts—creating a more equitable and efficient system without increasing the fleet size.

Load Balancing through Destination Flexibility

Finally, flexibility can extend to the destination itself. In "first-mile" transit, where the goal is to reach a transit network rather than a specific building, Siew Kei Lam and Guiyuan Jiang introduced the Peak-Hour Vehicle Routing Problem (PHVRP). By allowing a single request to be fulfilled at any of several nearby transit hubs, the routing algorithm can load-balance across the city, avoiding congested hubs and reducing total travel time for the entire fleet.

Whether it is walking 100 meters to a virtual stop, shifting a departure by ten minutes, or choosing a different subway station, these small concessions transform the transit problem from a rigid search for paths into a dynamic negotiation between agents. The result is a system that is not just faster, but more resilient and capable of serving significantly more people with the same number of vehicles.

Papers in this version

History & citationVersion 1 · 2026-10-04

Merged into another direction · Permanent link

View frozen version 1

Cite this direction

Latest version (a living page):

@misc{lacuna_optimizing_urban_transit_through_passeng_858c71d3,
  title = {Optimizing Urban Transit through Passenger Spatial and Temporal Flexibility},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/optimizing-urban-transit-through-passenger-spatial-and-temporal-flexibility/rdr_858c71d3019d48918a3ecda929a83451},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-19},
}

This exact version (version 1, frozen text):

@misc{lacuna_optimizing_urban_transit_through_passeng_858c71d3_v1,
  title = {Optimizing Urban Transit through Passenger Spatial and Temporal Flexibility},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/optimizing-urban-transit-through-passenger-spatial-and-temporal-flexibility/rdr_858c71d3019d48918a3ecda929a83451/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-19},
}