Grid Manufacturing : A Cyber-Physical Approach with Autonomous Products and Reconfigurable Manufacturing Machines

Each period of time has its own revolution, and with each revolution comes its own organisational model. We find ourselves in the 4th industrial revolution, where the internet of things connects autonomous embedded systems that live both in the virtual 'cyber' world, and in the real 'physical' world. These so-called Cyber-Physical Systems follow modern organisational models like self-management and can take proactive actions themselves. The thesis zooms in from the Cyber-Physical perspective to manufacturing systems that are both reconfigurable, autonomous, and flexible. This is achieved by developing new methods and using technologies that enable flexibility. However, efficiency has to be improved as well, e.g. by making assembly so flexible that it becomes cost-efficient to automate the production of low quantities of different products: so-called high-mix, low-volume production. The ability to automatically manufacture high-mix, low-volume production will drive the market for mass customisation, and bring about a shorter time to market. In practice, the move for flexibility will be achieved by creating new methods and tools, combining new technologies, and applied testing with simulators and newly-developed manufacturing systems. This thesis start by introducing the concept behind the manufacturing methodology, called 'Grid Manufacturing'. Grid Manufacturing takes place using autonomous entities, i.e. agents for both the equipment and the product itself. Products live in the 'cyber' world even before they are created in the 'physical' world and are aware of how they should be manufactured. They communicate and negotiate with reconfigurable manufacturing machines, called 'equiplets'. The equiplets offer generic services, e.g. pick & place, which a variety of products can use on demand by negotiating with the equiplets. This study focuses on the design and technology behind equiplets specifically and the infrastructure in general that is required to develop such a flexible and reconfigurable manufacturing system. To enable Grid Manufacturing, a whole set of technological challenges has been investigated. Where the main effort was put into the creation of automated flexibility for the manufacturing systems itself. This starts in the Chapter Object Awareness, which introduces an approach to dynamically handle and localise products by combining knowledge from different autonomous systems. The chapter is followed by Chapter Reconfiguration, which shows how products can communicate and control equiplets without being aware of the equiplet hardware configuration. The same chapter also shows how the hardware cannot just be used by a product, but can also be changed, making it possible to adapt the hardware configuration without reprogramming the system. Subsequently, in Chapter Architecture, the study focuses on combining performance and flexibility in a hybrid architecture that is used to control the 'grid' with equiplets, showing a hybrid platform made up of both a Multi-Agent System and the Robot Operating System platform for hardware control. After the architecture is established, the thesis focuses on how the architecture can be used safely, and introduces the control systems and system behaviour to be able to predict the behaviour of the equiplets. Now these fundamental systems for Grid Manufacturing have been covered it is time to continue with Chapter Validation and Utilisation, which looks from a grid perspective and shows how a grid can be used in new ways, with either a heterarchical control — in which all systems are equal — or not, when this is more efficient. The chapter also tests the grid with different simulated cases, to show that reconfiguration and the heterarchical approach can have several benefits