This paper proposes a methodology for modeling and controlling a flexible material handling system (MHS), composed of multiple automated guided vehicles (AGVs), suitable for Flexible Manufacturing Systems (FMSs). The AGVs incorporate artificial intelligent techniques to: i) facilitate the configuration and adaptation when there are layout modifications and ii) simplify the interaction between them using simple coordination models. In order to achieve higher flexibility, the MHS makes use of a decentralized navigation control, which increases autonomy and scalability, and a distributed Petri net for solving task allocation and traffic control problems. In order to facilitate the integration with the manufacturing processes, tasks dispatched by manufacturing cells are allocated by the MHS itself, taking into account pending transportation tasks and the system's performance. The whole system has been tested in a real factory and is currently in operational use.

This work addresses the path tracking problem of industrial guidance systems used by automated guided vehicles (AGVs) in load transfer operations. We focus on the control law that permits to AGVs to operate tracking a predefined route with industrial grade of accuracy, repeatability and reliability. One of the main issues of this problem is related to the important weight variation of AGVs when transporting a load, which induces slipping and skidding effects. Besides, localization error of the guidance system should be taken into account because position estimation is typically performed at a low sample rate. Other key point is that control law oscillations can knock down the load, which gives rise to safety and performance problems. Three control techniques - fuzzy, vector pursuit and flatness-based control - are compared in order to evaluate how they can deal with these problems and satisfy the robustness requirements of such an industrial application.

Purpose – The paper aims to describe the design and development of an automated guided vehicle (AGV) that incorporates artificial intelligence techniques to increase its autonomy and flexibility. The aim is developing a flexible AGV that operates as a flexible material handling system (MHS) in dynamic industrial environments. Design/methodology/approach – Introduces the entire on-board control system including hardware and software designs. The sensory system consists of a laser navigation system for localisation and a security laser scanner for sensing the environment. The software architecture is instantiated in a CPU that is connected to low level controllers through a CAN bus. Simplicity, flexibility, robustness and safety were concerned in the design process. Findings – The developed prototype is able to operate in partially structured and dynamic environments, is easily configured using an approximated description of the workplace and is able to adapt when slight floor layout modifications. This development shows that current technology permits introducing intelligent vehicles in complex manufacturing systems. Practical implications – The prototype is successfully tested in a real factory, operating as a flexible MHS, transporting pallets between production and storage lines. Originality/value – A novel flexible AGV is designed and developed to operate as a flexible MHS in dynamic industrial environments. The system satisfies the safety and robustness requirements of industrial applications. The flexible MHS results especially suitable for manufacturing systems that suffers from cyclic and seasonal variations and for flexible manufacturing systems where the possibility of choosing alternative routes is a must.

Efficient computation of density-based topology optimization using GPU computing is proposed.Workload of CUDA threads is properly balanced by the multi-granular GPU implementation of computationally intensive tasks.The use of on-chip memory is maximized in matrix-vector product operations using 3D CUDA kernels.The speedup and wall-clock time of the solving stage are evaluated using Jacobi and geometric multigrid preconditioning techniques.Numerical experiments show that the GPU instance of PCG using geometric multigrid preconditioner provides good performance results. This paper proposes a well-suited strategy for High Performance Computing (HPC) of density-based topology optimization using Graphics Processing Units (GPUs). Such a strategy takes advantage of Massively Parallel Processing (MPP) architectures to overcome the computationally demanding procedures of density-based topology design, both in terms of memory consumption and processing time. This is done exploiting data locality and minimizing both memory consumption and data transfers. The proposed GPU instance makes use of different granularities for the topology optimization pipeline, which are selected to properly balance the workload between the threads exploiting the parallelization potential of massively parallel architectures. The performance of the fine-grained GPU instance of the solving stage is evaluated using two preconditioning techniques. The proposal is also compared with the classical CPU implementation for diverse topology optimization problems, including stiffness maximization, heat sink design and compliant mechanism design.