e-Work based collaborative optimization approach for strategic logistic network design problem Pablo A. Miranda a , Rodrigo A. Garrido b , José A. Ceroni c, * a Network Configuration Research Group and School of Industrial Engineering, Pontificia Universidad Católica de Valparaíso, Av. Brasil 2241 – Casilla 4059, Valparaíso, Chile b Department of Transport Engineering, Pontificia Universidad Católica de Chile, Vicuña Mackenna 4860, Macul, Casilla 306, Santiago 22, Código 105, Santiago, Chile c Collaborative Systems Research Group and School of Industrial Engineering, Pontificia Universidad Católica de Valparaíso, Av. Brasil 2241 – Casilla 4059, Valparaíso, Chile article info Article history: Available online 31 December 2008 Keywords: e-Work Collaborative optimization Logistic distribution network design Inventory location problem Fleet design Customer clustering abstract We propose a collaborative e-Work based optimization approach for assisting the strategic logistic net- work design problem, which considers fleet design and customer clustering decisions. Normally, fleet design and customer clustering decisions are made by mid-level logistics managers while network design decision is made independently by high-level logistics managers. In the proposed approach, strategic dis- tribution network design is modeled as a Facility Location Problem, considering long term inventory con- trol decisions. On the other hand, tactical fleet design and customer clustering decisions are modeled based on a Hub and Spoke cost structure. An e-Work based heuristic approach is proposed to solve col- laboratively the network design problem at strategic and tactical levels. The collaborative solution approach results from a particular sequential decomposition of the problem, similarly to traditional loca- tion–allocation heuristics, modeling an information sharing strategy between decision makers involved at each organizational level. A numerical application of the approach with real data based instances shows significant benefits, when compared to a non-collaborative independent optimization, where the hierarchical levels share no dynamic information and base their decisions on static and independent optimization models. These results show evidence of the advantage of the e-collaborative approach to deal with logistic decisions at different hierarchical levels, organizational units, or companies, compared to non-integrated non-linear mixed integer programming models. Ó 2008 Elsevier Ltd. All rights reserved. 1. Introduction Supply chain management (SCM) usually deals with problems of different nature and at hierarchical levels as discussed in Coyle, Bardi, Langley (1992), Miranda (2004), Miranda and Garrido (2004a), Garrido (2001), Mourits and Evers (1995) and Simchi-Levi, Kaminsky, and Simchi-Levi (2003), Chopra and Meindl (2003), Da- vies and Spekman (2003), Bradley and Arntzen (1999) and other related literature. Traditionally and contrary to the approach pro- posed in this research, each problem in SCM has been tackled by different organizational units within the same or among compa- nies, based on independent and local methodologies. This sub-opti- mality of any independent approach shows the necessity for developing collaborative e-Work based methods. In terms of supply chain efficiency, the network’s ability to re- spond quickly and effectively to customers requirements is a key factor differentiating competing companies nowadays. Technolog- ical advancements through e-Work facilitate the way for meeting customer needs. Agility, quality, traceability, and service-ability depend on collaborative logistic networks that can focus on enhancing their performance by cooperative processes. Collabora- tive logistic networks are collections of supply chain partners designing and providing high quality and customized products (Chen, Chen, & Zhang, 1999). The general scope of e-Work was de- fined at the PRISM Center (Nof, 2000a, 2000b, 2003; Nof & Huang, 1998) as ‘‘any collaborative, computer-supported and communica- tion-enabled productive activities in highly distributed organiza- tions of humans and/or robots or autonomous systems”, as it can be found in nowadays supply networks. These collaboration pro- cess and modeling within e-Work have been defined as e-collabo- ration. Under the overload of tasks and information, collaboration must be enabled, and as much as possible optimized, among peo- ple, machines and computers. Without effective e-collaboration, the potential of emerging and promising electronic work activities, e-Work, cannot be fully materialized (Nof, 2003). The four main circles of knowledge in e-Work related literature are: ‘‘e-Work”, ‘‘Integration Coordination and Collaboration”, ‘‘Dis- tributed Decision Support” and ‘‘Active Middleware” (Nof, 2003). In the ‘‘Distributed Decision Support” circle, research is focused on interaction-based decision making models with distributed 0360-8352/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.cie.2008.12.007 * Corresponding author. Tel.: +56 32 273708; fax: +56 32 273806. E-mail addresses: pablo.miranda@ucv.cl (P.A. Miranda), rgarrido@ing.puc.cl (R.A. Garrido), jceroni@ucv.cl (J.A. Ceroni). Computers & Industrial Engineering 57 (2009) 3–13 Contents lists available at ScienceDirect Computers & Industrial Engineering journal homepage: www.elsevier.com/locate/caie