Technical and scientific objectives and innovative aspects

The general project objective is to develop and to implement an information and communication infrastructure to establish eBusiness standards for networking enterprises, based on agent technology (agents can be seen as software modules with intelligent and autonomous features).

MaBE will develop a multiagent business environment to enable dynamic and automated combination of multiple, complementary and alternative services between different users and providers along the whole value chain .

This enables internet-based and knowledge-driven discovery and selection of any business services, supporting service registration, resource allocation, negotiation, co-operation as well as automated collaboration processes and pre-selection of services within a net-economy.

MaBE also develops the connection of the common business environment to electronic marketplaces for business services. Such existing e-platforms extend the MaBE approach to an open business environment. The MaBE approach will serve as a decentralised environment for services that can dynamically be offered and ordered without reconfiguration.

Overview of scientific and technical workplan, main project output, milestones

To minimise the risk of this challenging project the structure of the workplan follows an iterative development process from industrial user requirements through multiagent system design to the final prototype implementation. The main project output is an agent-based software prototype for the decentralised discovery, selection and registration of business services. More detailed outputs are "Multi-sided Use Case Scenarios", "Agent System Concept", "1st Prototype" and "2nd Prototype", which integrate the research objectives and technical objectives (see also section 2). Due to the high complexity there is a Milestone after each major phase to evaluate the project progress and to be able to react on critical project states.

Research Objective


Objective


Description


RO1
Knowledge-based
Virtual Enterprise


RO1 consists of various knowledge-based co-operation and interaction scenarios for the networking Virtual Enterprise. These scenarios are the basis for the implementation of new efficient ways of co-operation of companies within Virtual Enterprises. (see TO 3)


RO2
Agent scenarios


RO2 consists of agent interaction and co-operation scenarios based on existing business processes of the consortium members. These scenarios are the basis for the implementation of the MaBE Multiagent business environment. (see TO1 -TO3)


RO3
Agent architecture


RO3 consists of an agent architecture supported by FIPA compliant platforms. This architecture allows each agent running in the MaBE business environment to communicate and interact with any other agent that supports FIPA compliant platforms without any modifications.


RO4
Agent
communication
infrastructure


RO4 consists of an agent communication infrastructure for enterprise collaboration on various levels of detail. Agents of retailers, service providers or other partners have to use this communication infrastructure for communication and interaction. Part of this communication infrastructure is an agent ontology (the vocabulary used by an agent). This ontology at least consists of the intersection of the vocabulary of the different service agents (e.g. logistic agent, retailer agent) to enable communication and interaction of different service agents. Each individual service agent extends the common ontology as much as necessary to fulfil the defined tasks.


Technical objectives


Objective


Description


TO1
Decision support


TO1 consist of the implementation of a software agent that pre-selects the "best" service provider out of a fixed (static) pool of providers. TO1 consists also of an interface to existing Internet-platforms or marketplaces that allows the pre-selection of alternative services and the collection of information. This pre-selection increases the speed of the decision making process. We are thinking of implementing agents playing business games using different business strategies. This allows the agent to collect knowledge that could later on be used in real business scenarios. Such business games will lead to a continuous improvement of the pre-selection done by agents used within the MaBE business environment.


TO2
Negotiation


TO2 consist of the implementation of software agents that are able to negotiate with each agent within the fixed pool of service providers. This allows automated negotiation e.g. for dynamic pricing. Such agents will increase the speed of partner negotiations (e.g. between retailers and logistic service providers).


TO3
Virtual Enterprise
knowledge-triggered agents


TO3 consists of implementation of knowledge-triggered software agents within the Virtual Enterprise. TO3 enables networking SMEs t o automate (partially) the resource allocation process within the knowledge intensive production process of a Virtual Enterprise.