文摘
The problem of resource-constrained scheduling of testing tasks for new product developmentis addressed. The problem is important because, in some industries like pharmaceutical andagrochemicals, a new product is required to pass all the tests by federal laws. If a product failsany of the tests, then all the remaining work on that product is halted and the investment inthe previous tests is wasted. Continuous time mixed-integer linear-programming models basedon two different respresentations of resource constraints are presented to solve this schedulingproblem. These models take into account complex trade-offs and have the capability of derivinga schedule that satisfies the resource constraints and utilizes the option of outsourcing. It isdemonstrated with the second model, which is based on graph representation and makes use oflogic, that the proper combination of modeling and search strategy can make the difference insuccessfully tackling this problem. Finally, it is also shown that it is critical to incorporateresource constraints along with sequencing of testing tasks to obtain a globally optimal solution.