The Concrete Delivery Problem

How to deliver, say, 600 vehicle loads of concrete to 200 construction sites with 150 concrete mixer vehicles in order to minimize the costs? When sending where the vehicles can make big differences in costs. Since concrete stays in workable shape only for a certain time inside a vehicle, such concrete delivery problems are subject to constraints, that cannot be found in traditional vehicle routing problems. In this thesis, these logistical problems are analyzed and several models developed. An algorithm based on large neighborhood search is presented, which is capable of solving big real-life instances of such concrete delivery problems heuristically. Although the thesis is mostly concerned with static optimization, issues of and extensions to dynamism and uncertainty are discussed.