TY - JOUR
T1 - A machine learning-driven two-phase metaheuristic for autonomous ridesharing operations
AU - Bongiovanni, Claudia
AU - Kaspi, Mor
AU - Cordeau, Jean François
AU - Geroliminis, Nikolas
N1 - Publisher Copyright:
© 2022 Elsevier Ltd
PY - 2022/9
Y1 - 2022/9
N2 - This paper contributes to the intersection of operations research and machine learning in the context of autonomous ridesharing. In this work, autonomous ridesharing operations are reproduced through an event-based simulation approach and are modeled as a sequence of static subproblems to be optimized. The optimization framework consists of a novel data-driven metaheuristic within a two phase approach. The first phase consists of a greedy insertion heuristic that assigns new online requests to vehicles. The second phase consists of a local-search based metaheuristic that iteratively revisits previously-made vehicle-trip assignments through intra- and inter-vehicle route exchanges. These exchanges are performed by selecting from a pool of destroy–repair operators using a machine learning approach that is trained offline on a large dataset composed of more than one and a half million examples of previously-solved autonomous ridesharing subproblems. Computational results are performed on multiple dynamic instances extracted from real ridesharing data published by Uber Technologies Inc. Results show that the proposed machine learning-based optimization approach outperforms benchmark state-of-the-art data-driven metaheuristics by up to about nine percent, on average. Managerial insights highlight the correlation between selected vehicle routing features and the performance of the metaheuristics in the context of autonomous ridesharing operations.
AB - This paper contributes to the intersection of operations research and machine learning in the context of autonomous ridesharing. In this work, autonomous ridesharing operations are reproduced through an event-based simulation approach and are modeled as a sequence of static subproblems to be optimized. The optimization framework consists of a novel data-driven metaheuristic within a two phase approach. The first phase consists of a greedy insertion heuristic that assigns new online requests to vehicles. The second phase consists of a local-search based metaheuristic that iteratively revisits previously-made vehicle-trip assignments through intra- and inter-vehicle route exchanges. These exchanges are performed by selecting from a pool of destroy–repair operators using a machine learning approach that is trained offline on a large dataset composed of more than one and a half million examples of previously-solved autonomous ridesharing subproblems. Computational results are performed on multiple dynamic instances extracted from real ridesharing data published by Uber Technologies Inc. Results show that the proposed machine learning-based optimization approach outperforms benchmark state-of-the-art data-driven metaheuristics by up to about nine percent, on average. Managerial insights highlight the correlation between selected vehicle routing features and the performance of the metaheuristics in the context of autonomous ridesharing operations.
KW - Dial-a-ride problem
KW - Electric autonomous vehicles
KW - Large neighborhood search
KW - Machine learning
KW - Metaheuristics
KW - Online optimization
UR - http://www.scopus.com/inward/record.url?scp=85136188486&partnerID=8YFLogxK
U2 - 10.1016/j.tre.2022.102835
DO - 10.1016/j.tre.2022.102835
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
AN - SCOPUS:85136188486
SN - 1366-5545
VL - 165
JO - Transportation Research Part E: Logistics and Transportation Review
JF - Transportation Research Part E: Logistics and Transportation Review
M1 - 102835
ER -