Hands-on adoption of a multi-agent production planning technology in the manufacturing industry.
ExPlanTech is part of the trial projects of the Information Society Technologies (IST) Programme for Research and Development of the European Commission, which aims at the adaptation and introduction of (promising but not yet fully established) leading edge technology in industrial/service applications and its evaluation by both supplier and user.
The objective of the ExPlanTech trial is to introduce and customise a novel multi-agent production planning technology (ProPlanT system) in two specific industrial cases in Greece and Czech Republic. Our aim in the project is to demonstrate a new quality of project-driven production planning based on multi-agent engineering and distributed processing as compared to the traditional solutions.
The new multi-agent technology offers opportunities to adapt industrial manufacturing to more demanding customer needs while optimizing resource utilisation, intelligent decision processes and supplier chains. Moreover, our aim is to integrate the customized ProPlanT prototype into existing Enterprise Resource Planning (ERP) systems at real end users sites, as well as to validate the new approach of production planning and demonstrate an enhanced performance in connection with end user ERP systems.
About the Technology
ProPlanT technology addresses the problem of project-driven production with emphasis on specific customer requirements and intelligent decision support in order to improve supplier/consumer relationships, working environment and customer confidence.
Traditional production planning activity, as still approached in the majority of industrial enterprises, is substituted by an agent-driven service negotiation and decision process based on usage-centred knowledge about task requirements. The new technology increases business and technological opportunities in the global IT market and extends the competitiveness of SMEs by optimizing their operational capabilities.
The project work consists of the following phases:
-- Adapting and customizing the ProPlanT technology platform, including communication interfaces, towards two end user specification needs.
-- Encapsulating and installing the customised prototype of the multi-agent production planning system within the existing ERP systems in the two specific industrial end user sites and testing the functionality with respect to stability of the new system.
-- Conducting and analysing a measurement and evaluation exercise on real industrial data at the two end user sites. Such an analysis forms the basis for the dissemination and exploitation activities, which will be an essential part of the project.
The evaluation and exploitation process will be carried out on a long-term basis and continue after the project lifetime.
The Work Plan
As stated before, the mission of the ExPlanTech Trial project is to introduce, customise and exploit a novel multi-agent production planning technology in two specific industrial cases. Our aim is to demonstrate an enhanced quality of project-driven production planning based on agent-based engineering in connection with the Enterprise Resource Planning System (ERP) and to help facilitate its wider usage within an increased number of industrial applications.
It is generally accepted that production planning is a complex industrial issue, the automation of which is inevitable. In the context of the trial, the ProPlanT technology is provided in the form of a prototype of an integrated multi-agent system for automation of customer-driven production planning and administration. ProPlanT has been developed within the framework of the European collaborative research project EUREKA no.: 1439 - PVS '98 led by the Gerstner Laboratory, Czech Technical University in Prague (CTU FEE), Czech Republic, which acts in the ExPlanTech project as the technology supplier partner.
The Gerstner Laboratory (CTU FEE) successfully substituted classical planning and scheduling mechanisms with the process of negotiation, job delegation and task decomposition within a community of autonomous agents, each of which represents production or information unit of the modelled factory.
The ExPlanTech Business Case
The project aims at presenting a new quality of project-driven production based on multi-agent technology as compared to the traditional planning solutions. The industrial cases solved in this trial will show the benefits of this approach -- e.g. improved resource allocation and flexibility, better supplier/customer relationships in terms of end-users.
Though the original laboratory prototype was tested on generalised "toy problems," resembling fundamental features of production planning domain, the methodology needs to be exposed and evaluated in real industrial enterprise environments. The software result of the trial is expected to be a fully functional instance of user centred production planning multi-agent system tailored to the needs of the ExPlanTech Trial end-user partners.
There are two end-user manufacturing partners in the consortium. Liaz Pattern Shop Ltd. is our Czech manufacturing partner, who belongs to leading European producers of patterns and forms for the automobile industry (mainly VolksWagen group). Hatzopoulos S.A., has been the major Greek player in the business of producing flexible packaging materials since 1931, working mainly for food companies including multinationals such as Nestle, Pepsico, Warner Lambert, Procter & Gable, etc.
Overall Positioning of the Project
The ExPlanTech combined technology adaptation and introduction successfully addresses the problem of industrial enterprises (mainly SMEs) facing increasing varieties of offered products together with decreasing volumes of production and increased customer needs orientation.
Ongoing production planning activities are required to meet specific customer requests. But the current planning process is rather re-active and incidental rather than pro-active and strategy-driven. Both sales and R&D staff are busy with daily operational activities.
As a result, new product and market opportunities are not realised and taken, and product life-cycles are becoming over-aged, technological opportunities get missed and the customer-oriented product development is translated to the single-customer product development. Further consequences can be illustrated with potential turnover and high internal overheads, material and logistics costs caused by over-aged and partly over-customised products.
Thanks to the ProPlanT multi-agent technology, the company assortment will be aligned with existing as well as future customer needs. This will improve supplier/consumer relationships and will result in increased productivity, resource efficiency, improved working environment and consumer confidence. This will directly contribute to enhance competitiveness for the technology/end-users and increase new business and technological opportunities in the global IT market.
Adaptations at the level of the two specific end-user industrial environments are supported through both adaptable design of the agents' community and run-time adaptability to the particular needs of the collaborating IT actors. Adaptability focuses on the notion of supporting the automatic provision of various "interface" versions with different production planning characteristics, based on usage-centred knowledge about customer needs, abilities and preferences as well as task requirements.
Hence, the resulting production planning environment can present various types of usage scenarios to different user categories in an automated fashion. By system's adaptability we mean the ability of interactive production planning multi-agent applications to dynamically modify the context attributes, communication scripts and community architecture resulting in changing the system's behaviour with respect to efficiency improvements, optimal utilisation of resources, etc.
Finally, the trial helps to extend the accessibility of SMEs towards new promising (but not fully established) technologies in order to conduct business to the increased number of users while being more adaptable to the customer's needs.
Adamantios Koumpis heads the Research Programmes Division at Unisoft S.A., Greece. He is coordinator of the Explantech Project funded by the European Commission. His research interests include quantitative decision making techniques and Information Society economics. His e-mail is: [email protected].