A problem design and constraint modelling approach for collaborative assembly line planning. (February 2019)

• Record Type: Journal Article
• Authors: Neumann, Daniel; Keidel, Jan
• Is Part Of: Robotics and computer-integrated manufacturing. Volume 55(2019)Part B
• Page Start: 199
• Page End: 207
• Publication Date: 2019-02
• Subjects: Collaboration -- Assembly line design -- Assembly line planning -- Collaborative planning -- Collaborative problem modelling; Robots, Industrial -- Periodicals; Computer integrated manufacturing systems -- Periodicals; Robotics -- Periodicals; Robots industriels -- Périodiques; Productique -- Périodiques; Robotique -- Périodiques; 670.285
• DOI: 10.1016/j.rcim.2018.05.002
• Languages: English
• ISSNs: 0736-5845

Abstract

Highlights: An adaptable multi-agent systems has to allow the modelling of the dependencies between relevant elements of the assembly line planning. Virtual collaboration for modelling problem model relevant elements, objectives and constraints, requires a robust conflict search. The acceptance of the solutions only corresponds to the optimal solution when the problem is described perfectly. The perfect collective problem description increases the graph complexity and outsources the solution search to the problem definition. The solution search speed using genetic algorithm is shown to be dependent on the graph complexity and the graph flow. Abstract: Assembly line planning is the interface between agents of different professions, competencies, skills and experiences. Each agent is characterized by specific objectives and constraints, that must be considered when different products and processes are merged across multiple levels of various decision making units. A systemic platform for the dependencies of goals, constraining relationships of elements to be assigned and the interdependencies to properties of the elements is designed. Hereby, the objectives and the constraints are evaluated and mapped within the assembly line planning (ALP) process, resulting in a comprehensive graph with the processes, structures, resources and their properties with its dependencies. The dependencies include time criteria but also technical, processual, ergonomic, qualitative, part-dependent along with resource-dependent, structural constraints. The dependency map is integrated within an existing ALP process to achieve a temporal precedence on the incoming and outgoing information resulting in a dynamic process oriented presentation of the objectives and constraints. The elements of the dependency map are categorized and analysed based on the evaluations and a problem model is created for the presented ALP problem. The problem model is solved using a genetic algorithm which is parameterized in accordance to the problem model complexity and characteristic.