Solving Truckload Procurement Auctions Over an Exponential Number of Bundles

In the United States, hundreds of billions of dollars worth of services to shippers are provided by truckload carriers each year. Shippers are provided with a fast and easy way to negotiate potential contracts with a large number of carriers through Internet auctions. Allowing multiple lanes to be considered simultaneously in a single auction is an added benefit of combinatorial auction. This is important because carriers are enabled to connect multiple lanes in continuous moves or tours, which decreases empty mileage to be driven, and therefore increases overall efficiency. Combinatorial auctions, on the other hand, require bidding on an exponential number of bundles if full economies of scope and scale are to be achieved, which is only tractable in very small auctions. Bidding instead typically is limited to a very small subset of potential bids in most real-world auctions. The authors present an implicit bidding approach to combinatorial truckload procurement auctions that enable implicit consideration of the complete set of all possible bids, with out placing on the bidders or the auctioneer the corresponding burden of an exponential number of bids. The authors present the needed models to solve this problem. The authors the demonstrate the tractability of their approach by providing extensive computation results. Finally, the authors conclude with numerical analysis to assess quality of the generated solution and demonstrate the benefits of their approach over existing bidding methods in practice.

  • Availability:
  • Authors:
    • Chen, Richard Li-Yang
    • Beygi, Shervin Ahmad
    • Cohn, Amy
    • Beil, Damian R
    • Sinha, Amitabh
  • Publication Date: 2009-11


  • English

Media Info

  • Media Type: Print
  • Features: Appendices; Figures; References; Tables;
  • Pagination: pp 493-510
  • Serial:

Subject/Index Terms

Filing Info

  • Accession Number: 01149176
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Jan 22 2010 2:43PM