An agent-based approach to engineering design Kuo-Ming Chao a,* , Peter Norman b , Rachid Anane a , Anne James a a School of Mathematical & Information Sciences, Coventry University, Priory Street, Coventry CV1 5FB, UK b EngineeringDesign Centre, University of Newcastle-upon-Tyne, Armstrong Building, Newcastle-upon-Tyne, UK Abstract Among the features of concurrent engineering is the notion of distributed design, and the ability to communicate design changes to multidisciplinary teams. Engineering design is a complex activity. Differences in system architectures and information structures, and co-ordination requirements tend to reduce the effectiveness of distributed design. Current thinking indicates that multi-agent systems MAS) can alleviate some of the complex engineering design problems. In this paper, it is argued that agent attributes such as proactiveness and autonomy can overcome these limitations. Agents provide a ¯exible and dynamic approach to distributed/multidisciplinary design team which can reduce redundant design activities, and improve co- ordination. # 2002 Elsevier Science B.V. All rights reserved. Keywords: Intelligent agent; Proactiveness; ORB; Engineering design 1. Introduction The successful design of large and complex made- to-order MTO) products such as ships, offshore oil platforms and aeroplanes, require the collaboration of multidisciplinary design teams. These teams use spe- cialised computer systems to aid their design process. Unfortunately, such computer systems may use dif- ferentdatarepresentationsoftheproductmodel.They may also utilise different design software packages. These packages may be written in dissimilar lan- guages,forinstanceC,C,Javaorotherlanguages, and installed on different hardware systems. Conse- quently, any collaborative communication or co-ordi- nation between such diverse and different models, languages and system architectures may prove dif®- cult. Within the last decade, a number of design tools have been developed at the Engineering Design Cen- tre, University of Newcastle. The following brie¯y describes these:  Process flow diagram PFD) system [1], used to design a gas-condensate separation process. The system generated a PFD frame and rule model of the process.  Electrical system [2], computes the power demand and power generation equipment dimensions. The system can also select equipment from the suppli- er's catalogues. This selection is based on the PFD specifications.  Plant operation system [2], produces diagnostic rules for functioning plant. These rules are based on the output of the PFD system.  Associativity data generation ADG) [3], calculates the strength of the relationship between any two pieces of equipment based on connectivity, func- tion,andcost.Theequipmentandconnectivitydata Computers in Industry 48 2002) 17±27 * Corresponding author. Tel.: 44-24-76888908. E-mail address: k.chao@coventry.ac.uk K.-M. Chao). 0166-3615/02/$ ± see front matter # 2002 Elsevier Science B.V. All rights reserved. PII:S0166-361502)00007-6