Reconfigurable robotics for responsive manufacture - R3M

To be truly responsive, a manufacturing system should be able to rapidly adapt to what the production need is at a specific time, depending on demand rather than on capacity. Conventional automation cells tend be fixed and specifically designed to manufacture a single or limited number of products in large volumes. However, where product volumes or types are highly variable this approach is inefficient and costly. More resilient approaches that can rapidly adapt to variations in product quantity and type are of great interest and a significant quantity of work has been carried out to realise this concept. However, whilst physical reconfiguration, ie. the positioning of robots and process systems, is relatively easy to achieve, the major barrier is the need for time consuming and costly reprogramming to support each change. The research we propose will take a more holistic view of the reconfiguration process and develop new algorithms that can automatically generate programme and configuration data from CAD and process data eliminating the need for significant human input. Furthermore the system will also consider safety and how to automatically configure the safety system so that it is safe and legally compliant but also implement a flexible framework that allows the active intervention of human operators. Within the research we bring together experts in Robotics, AI and Control and Automation from three leading Universities to work together and develop game changing approaches to resilience in manufacturing. We will also engage with a number of end users and suppliers to ensure that the developed science has real world relevance and is aligned with realistic industrial challenges.