Efficient Auctions for Distributed Transportation Procurement

Su Xiu Xu,George Q. Huang
DOI: https://doi.org/10.1016/j.trb.2014.03.005
2014-01-01
Abstract:The purpose of this paper is to propose allocatively efficient auction mechanisms for the distributed transportation procurement problem (DTPP), which is generally the problem of matching demands and supplies over a transportation network. We first construct a one-sided Vickrey-Clarke-Groves (O-VCG) combinatorial auction for the DTPP where carriers are allowed to bid on bundles of lanes. The O-VCG auction minimizes the total transportation cost (i.e., allocative efficiency) and induces truthful bidding from carriers (i.e., incentive compatibility). To simplify the execution of auction, we next propose a primal-dual Vickrey (PDV) auction based on insights from the known Ausubel auctions and the primal-dual algorithm. The PDV auction is actually a multi-round descending auction that seems simple enough for bidders. The PDV auction realizes VCG payments and truthful bidding under the condition of seller-submodularity, which implies that the effect of each individual carrier is decreasing when the coalition increases. Such is the case for the DTPP in an oversupplied transportation market. The winner determination problem of O-VCG auction is solved by the proposed primal-dual algorithm when seller-submodularity holds. Finally, carriers may reveal less private information in the PDV auction due to its dynamic procedures. (C) 2014 Elsevier Ltd. All rights reserved.
What problem does this paper attempt to address?