A branch‐and‐bound algorithm for scheduling of new product development projects

Abstract

In this paper, we consider scheduling problem in a new product development project. Each research and development project consists of a set of activities in which each activity may be executed in several ways. Each way of execution of an activity is a different technology, called an alternative technology, which may fail due to technical risks. In this work, we focus on alternative technologies for scheduling activities to maximize the expected net present value. We assume that the activity payoff is obtained as soon as any single technology is completed successfully. Therefore, we analyze the problem and develop a two‐phase solution procedure, consisting of a branch‐and‐bound algorithm and a recursive search procedure. The efficiency of the proposed algorithm has been evaluated on a set of randomly generated test instances.

A Genetic Algorithm for Scheduling Alternative Tasks Subject to Technical Failure

“…Some alternatives may imply the implementation of other alternatives. This is a recently proposed problem and it has been addressed by exact methods only [22,23]. Since this is a NP-hard problem (see [23]), an exact algorithm without an exponential time complexity is unlikely to exist.…”

Exact mathematical formulations and metaheuristic algorithms for production cost minimization: a case study of the cable industry

“…A well‐known exact approach is branch and bound's method that can be applied by heuristic algorithms or optimization software, for example Cplex, Xpress, Lingo, and so on. This method was applied by Rios‐Mercado and Bard (), Rabadi et al (), Dunstall and Wirth (), Ranjbar (), Neto et al (), and so on. Different metaheuristic algorithms, for example simulated annealing (SA), genetic algorithm, tabu search, ant colony, and so on, were proposed as nonexact methods in order to obtain a good feasible solution where the degree of difficulty of the model is high and exact methods are not able to solve the mathematical model in reasonable time.…”