PLAN A – In a research cooperation with Fraunhofer IAO flexis develops an agent-based planning for rework and maintenance
Agent-based planning with uncertain and changeable order content
For companies a very strong focus is on reworking, since the products are largely completed, and thus very close to the customer and the realizable revenue. Similar processes regarding organization and content can be found in the maintenance, for example of engines, trains or airplanes. Their downtime has to be minimized for economic reasons.
Both, rework and maintenance are characterized by uncertainties regarding the content to be processed. The symptom “Engine does not start", which could be recognized upon the first inspection of a vehicle, may have various causes. For example, an unconnected battery or an empty fuel tank. Every cause must be resolved in a different way – with different operations, staff and resources. Moreover, the cause is often detected only after a detailed examination and disassembly. This results in a planning uncertainty, that may have an impact on subsequent orders, for instance if they have to pass through the same facilities.
Such uncertainty concerning the necessary effort creates a need for a flexible planning that responds quickly and efficiently to changes in the conditions and the order content. The guiding principle of PLAN A is to enable optimal planning for job-shop-manufacturing by making use of an agent system. This agent system consists of several identical or differently specialized acting units, collectively solving the problem. At PLAN A agents represent job content and negotiate the succession of order execution based on the current situation at the shop floor and the differing priority of the jobs.
As an application partner flexis was able to win over MTU Maintenance Berlin-Brandenburg GmbH, the MTU center of expertise for the maintenance of industrial gas turbines.
PLAN A is funded by the Federal Ministry of Education and Research as part of the "SME innovation campaign ICT".