The Optimal Resource Allocation in Stochastic Activity Networks via The Electromagnetism Approach Anabela P. Tereso, M. Madalena T. Araújo Universidade do Minho, 4800-058 Guimarães PORTUGAL anabelat@dps.uminho.pt; mmaraujo@dps.uminho.pt Salah E. Elmaghraby North Carolina State University, Raleigh, NC 27695-7906 USA elmaghra@eos.ncsu.edu October, 2004 Abstract An optimal resource allocation approach to stochastic multimodal projects had been previously developed by applying a Dynamic Program- ming model [2] [3], which proved to be very demanding computationally. Approximations to the initial model had been also developed, still within DP framework [4]. Computing times were improved, but demonstrated the need for further developments. In this paper we report on the appli- cation of a recently developed technique for global optimization, the Elec- tromagnetism Algorithm (EMA) [1], to this problem and demonstrate its superior performance to previously attempted approximations. Key Words: Project Management and Scheduling, Stochastic Activity Networks, Resource Allocation, Electromagnetism Algorithm. 1 Problem Definition The problem treated in this paper may be stated as follows. We are given a multimodal activity network (that is, each activity can be performed at any number of levels of resource intensity applied to it, with resulting shorter or longer duration), with stochastic work content {W a } for each activity a ∈ A, where A is the set of activities, with |A| = n. We assume that the W a ’s are 1