Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Optimizing Ride-Pooling Operations with Extended Pickup and Drop-Off Flexibility

Hao Jiang, Yixing Xu, Pradeep Varakantham

TL;DR

This work tackles city-scale ride-pooling by relaxing the conventional fixed pickup/drop-off assumption through flexible passenger walking to nearby points. The framework, FlexiPool, combines a tree-based feasible combination generator, a Regional Vehicle Routing Problem MILP for route optimization, and Neural Approximate Dynamic Programming with value decomposition to account for long-term rewards. Empirical results on a large taxi dataset show up to 13% more served requests and up to 21% shorter average travel distances compared with strong baselines, highlighting the operational gains from leveraging passenger mobility. The approach offers a scalable, forward-looking solution that could significantly improve efficiency and sustainability in urban ride-pooling systems, with future work including incentive-based pricing to encourage walking to meeting points.

Abstract

The Ride-Pool Matching Problem (RMP) is central to on-demand ride-pooling services, where vehicles must be matched with multiple requests while adhering to service constraints such as pickup delays, detour limits, and vehicle capacity. Most existing RMP solutions assume passengers are picked up and dropped off at their original locations, neglecting the potential for passengers to walk to nearby spots to meet vehicles. This assumption restricts the optimization potential in ride-pooling operations. In this paper, we propose a novel matching method that incorporates extended pickup and drop-off areas for passengers. We first design a tree-based approach to efficiently generate feasible matches between passengers and vehicles. Next, we optimize vehicle routes to cover all designated pickup and drop-off locations while minimizing total travel distance. Finally, we employ dynamic assignment strategies to achieve optimal matching outcomes. Experiments on city-scale taxi datasets demonstrate that our method improves the number of served requests by up to 13\% and average travel distance by up to 21\% compared to leading existing solutions, underscoring the potential of leveraging passenger mobility to significantly enhance ride-pooling service efficiency.

Optimizing Ride-Pooling Operations with Extended Pickup and Drop-Off Flexibility

TL;DR

This work tackles city-scale ride-pooling by relaxing the conventional fixed pickup/drop-off assumption through flexible passenger walking to nearby points. The framework, FlexiPool, combines a tree-based feasible combination generator, a Regional Vehicle Routing Problem MILP for route optimization, and Neural Approximate Dynamic Programming with value decomposition to account for long-term rewards. Empirical results on a large taxi dataset show up to 13% more served requests and up to 21% shorter average travel distances compared with strong baselines, highlighting the operational gains from leveraging passenger mobility. The approach offers a scalable, forward-looking solution that could significantly improve efficiency and sustainability in urban ride-pooling systems, with future work including incentive-based pricing to encourage walking to meeting points.

Abstract

The Ride-Pool Matching Problem (RMP) is central to on-demand ride-pooling services, where vehicles must be matched with multiple requests while adhering to service constraints such as pickup delays, detour limits, and vehicle capacity. Most existing RMP solutions assume passengers are picked up and dropped off at their original locations, neglecting the potential for passengers to walk to nearby spots to meet vehicles. This assumption restricts the optimization potential in ride-pooling operations. In this paper, we propose a novel matching method that incorporates extended pickup and drop-off areas for passengers. We first design a tree-based approach to efficiently generate feasible matches between passengers and vehicles. Next, we optimize vehicle routes to cover all designated pickup and drop-off locations while minimizing total travel distance. Finally, we employ dynamic assignment strategies to achieve optimal matching outcomes. Experiments on city-scale taxi datasets demonstrate that our method improves the number of served requests by up to 13\% and average travel distance by up to 21\% compared to leading existing solutions, underscoring the potential of leveraging passenger mobility to significantly enhance ride-pooling service efficiency.

Paper Structure

This paper contains 13 sections, 15 equations, 12 figures, 3 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Related Work
  3. Ride-pool Matching Problem(RMP)
  4. Overview of FlexiPool
  5. Extension of Pickup/Drop-off Area
  6. Feasible Combination Generation
  7. Route Optimization
  8. Future aware Matching
  9. Experiment
  10. Setup
  11. Experimental results
  12. Ablation Study
  13. Conclusion

Figures (12)

  • Figure 3: In the original setting (left), vehicle $v$ must travel to the passenger's original pickup location $p$. With flexible pickup and drop-off (right), passengers can walk to an alternative pickup point $p'$ that is more accessible for the vehicle.
  • Figure 4: Example of Combination Generation. Black solid lines represent feasible combinations, while black dashed lines represent infeasible combinations.
  • Figure : I. Collect Request from Road Network
  • Figure : A. Solution without flexible pickup and drop-off
  • Figure : I. Collect Request from Road Network
  • ...and 7 more figures