Building the Next Generation of Factories
Be a part of the Project

Dynamic Manufacturing

Real Time Dynamic Manufacturing

Current limitation in industry

Previously, production elements (e.g. humans, robots and machines) were “trained” at the commissioning of a factory to produce a specific product. As new design iterations evolved, significant resources were invested to retrain and restructure facilities for each variation in a step change fashion. This results in production delays that restricts the execution of new product lines to longer periods and increases the cost of variant production.

Additionally, current production methods are largely limited to the use of human or automated/robotic systems working in separate manufacturing cells. Human only production suffers from lack of productivity as humans are required to complete highly repetitive tasks that limit their ability to effectively operate. Robot only production is limited in its capacity to adjust to product design changes and requires massive overheads and specialist technical knowledge to implement change.

Current limitation in scope of research

Recent research has identified human-robot collaboration as a solution to increasing adaptability, and therefore productivity, in the assembly line by combining the benefits of a human worker in dexterity, intelligence and adaptability with the strength, optimized repeatability and accuracy of a robot. The further development of collaborative robots has enabled safe operation in direct contact with human workers to also increase productivity. The challenge now is to provide an environment where employees can effectively and efficiently operate with their robot companions inside the same environment.

The Value added to research

Chatty Factories therefore focuses on human-robot collaborative manufacture to provide a deep understanding of the way an expert worker performs with, and communicates tasks to, robotic collaborators. The proposed combination of advanced machine learning, human-machine pedagogical theory, interpretable data analytics and direct demonstration will enable robots and humans to make better decisions in a complex world and produce product variants “right the first time”.

The Value Added to Industry

Enhanced with advanced machine learning techniques informed by human-machine pedagogy, industrial production will move beyond staged manual refinement to dynamically adapt to product and production variances.

This will lead to significant industrial value by:
•Enabling dynamic reskilling of production elements on the factory floor to minimize training time and ensure maximum flexibility and efficiency when fabricating products that are varying in their design. The final production ecosystem will also account for different humans working with, teaching to and learning from, robotic elements.
•Improved machine learning capabilities based on reinforced demonstration and interpretable data analytics to increase production robustness.
•Looking beyond individual production elements to the full production process by creating a dynamic production plan that obeys regulatory restrictions while adapting to production variations. This will start with a master plan automatically generated during the design stage, then continuously updated during production to enable a fluidly evolving production environment.

Visualising the future

Results will enable manufacturing ecosystems that can continuously reskill and reorient the human and robot production elements to improve efficiency, robustness and compliance within any regulatory structure. This will provide a new model for fluid factory production that completely rethinks product use, design and manufacture through the use of data and intelligent algorithms to drive innovation and efficiency. Thus shattering the boundaries between the plant and the outside world, while retaining security and integrity.

Research Investigators:

Dr David Branson III and Dr Mojtaba Ahmadieh

Comments are closed.