An operational approach to automated tanker scheduling by computer

Abstract

That tanker scheduling problem considered in this paper is that of the Defense Fuel Supply Center (DFSC) and the Military Sealift Command (MSC) in the worldwide distribution of bulk petroleum products. Every six months DFSC receives from the Armed Forces estimates of quantities of products needed and dates during which deliveries should be made. Using these estimates DFSC contracts for the required products with various suppliers, and then works with MSC to construct cargoes and assign tankers to meet the requirement demands. Approximately 14 products, 125 ports, and 50 tankers are involved. A mathematical programming model (mixed integer) is developed which attempts to reflect the true scheduling task of DFSC and MSC as closely as possible. The problem is kept to within a workable size by sequentially focusing attention on requirements due to delivery during successive time periods, and by the systematic construction of a set of tanker routes which does not contain many routes that can be judged unacceptable from practical considerations alone.