A greedy approach for increased vehicle utilization in ridesharing networks

TL;DR

The paper tackles eco-friendly ridesharing by leveraging predicted origins and destinations to maximize multi-passenger routes under detour and capacity constraints. It introduces a k-hop sliding-window heuristic that dramatically reduces the NP-hard search space and proves the objective is monotone submodular, making greedy optimization effective. By transforming the problem into a vehicle-count framework and using a minimum path cover, it also derives an estimate of the optimal fleet size. Empirical results on NY and DC datasets show improved vehicle utilization, higher sharing rates, and lower fleet requirements, with real-time performance around 1.1 seconds per query. Overall, the approach offers a scalable, data-driven method to green-purpose routing in urban ridesharing networks.

Abstract

In recent years, ridesharing platforms have become a prominent mode of transportation for the residents of urban areas. As a fundamental problem, route recommendation for these platforms is vital for their sustenance. The works done in this direction have recommended routes with higher passenger demand. Despite the existing works, statistics have suggested that these services cause increased greenhouse emissions compared to private vehicles as they roam around in search of riders. This analysis provides finer details regarding the functionality of ridesharing systems and it reveals that in the face of their boom, they have not utilized the vehicle capacity efficiently. We propose to overcome the above limitations and recommend routes that will fetch multiple passengers simultaneously which will result in increased vehicle utilization and thereby decrease the effect of these systems on the environment. As route recommendation is NP-hard, we propose a k-hop-based sliding window approximation algorithm that reduces the search space from entire road network to a window. We further demonstrate that maximizing expected demand is submodular and greedy algorithms can be used to optimize our objective function within a window. We evaluate our proposed model on real-world datasets and experimental results demonstrate superior performance by our proposed model.

Figures (20)

• Figure 1: Road network represented in the form of a grid
• Figure 2: A particular instance of the road graph $G^R$. In this graph, the vertices represent the grid cells and the edge weight represents the distance between grid cells. The distance is measured between the center points of grid cells.
• Figure 3: In these graphs, the vertices represent grid cells, edges represent the direction of requests, and the edge weight represents the (a) predicted and (b) actual number of requests between the grid cells. For simplicity, we have ignored the edges with $0$ weight.
• Figure 4: A particular instance of the road network with the driver and passenger orders. The nodes represent grid cells and the edge weight represents the distance between the center points of grid cells. The vehicle is at grid cell $g_1$ and servicing the passenger who has to travel from $g_1$ to $g_4$. Meanwhile, the requests for two orders arrive at $g_6$ and $g_7$, and they have to reach $g_4$.
• Figure 5: Performance of demand and origin-destination prediction. In this graph, nodes represent grid cells and the node weight represents the expected number of passengers on the grid cell. The origin and destination of passenger $i$ are denoted as $o_{is}$ and $o_{id}$ respectively.