Lean, Green, and Agile Design and Control of Pallet Systems

Licentiate thesis, 2011

Pallet based automated material handling systems are main devices to transport and locate various part types in flexible production plants. To fulfill production objectives and win in the competitive environment of mass customization, proper deign and control of handling and locating pallet systems have been highly concerned. Minimization of energy, waste, and time are important keys in a pallet system design according to the lean, green, and agile philosophies. This thesis work focuses on these objectives and establishes a framework to design and control a pallet system according to the mentioned philosophies. In this framework, evaluation and synthesis methods are combined to maximize precision and minimize time of the design stages. For the evaluation method, a pallet system universal model to evaluate a wide range of possible design architectures and control strategies is introduced and developed. The great number of configurable parameters in this model provides a precise insight on the system analysis in transition and steady states. However, the precise analysis is often paid by massive evaluation time. To reduce the time, furthermore, mixed integer linear programming models are employed to synthesize optimal initial values of the configurable parameters according to some design assumptions. A systematic approach to construct the linear mathematical constraints characterizing the universal model in steady states is established. Moreover, mathematical models of the lean structure and the drive unit energy consumption are derived and used as design objectives in the optimization models. Both optimization and universal models are devised in a case study where a pallet system is designed based on a specific technology called X85. The results of this study indicate the crucial impact of the framework on lean, green, and agile design of pallet systems. This framework is also shown to be enhanced by resolving the time complexity, which may occur in the optimization models due to the binary characteristics of some decision variables and objective functions nonlinearity.