In combinatorial auctions, solving the bid generation problem (BGP) for bidders is critical to achieving efficiency. However, in the recent surge of combinatorial auction research, little attention has been paid to the BGP. In this paper, the BGP faced by transportation service providers is studied. First, the bidder's optimality criterion of a combinatorial bid is clarified, and then the focus is on the bundling method when an OR bid language is used. Through examples, bundles generated by solving the optimal truck routing problem were examined, and it was found that the resulting bid might not be optimal. This heuristic is compared with a simple nearest insertion method. The simulation result shows that whereas the former outperforms the latter in most cases, many times the latter outperforms the former by 5% to 8%.