A hybrid-agent-based approach for block break definition using fuzzy logic

This paper presents a hybrid-agent-based methodology for defining block breaks using fuzzy logic mechanisms for agent preference representation and mediation. The concept of a hybrid agent is defined, and the fuzzy-logic-based mechanisms for defining the preferences of individual agents and then deriving joint preferences are described. Working examples are given for defining blocks and associated breaks in the hold area and engine room of the notional fast sealift ship PD-337. The agents jointly defining the block breaks in these examples are a structure manufacturing planner, a structure assembly planner, and an outfitting planner. Excel spreadsheets are used in these examples to execute the fuzzy-based operations. This methodology has proven to be straightforward, to provide useful results, and to have the potential to improve the efficiency and optimality of block definition. The methodology could be effectively integrated into the detail/production design process and could also be applied to the definition of other planning units, such as outfit units and on-board outfitting zones. Areas of further research include exploring within the hybrid agent and fuzzy-based context (1) the ability to define breaks in all three dimensions simultaneously rather than sequentially where simultaneous three-dimensional optimization is necessary, (2) dynamic break definition where a break location is dependent to some extent on the location of adjacent breaks, (3) the simultaneous consideration of preferences from lifting/weight and planning/work content agents that use block attributes other than location to determine if breaks are satisfactory, and (4) the development of an integrated and user-friendly tool for block break definition for use in the production planning or build strategy development process.