It is widely acknowledged that AI is set to play a pivotal role for manufacturers in boosting efficiencies in the design and production stages, yet simulation seems to take a back-seat in the discussion about the future of manufacturing. In this article, Chief Technology Officer at CFMS, Ian Risk, calls on manufacturers to embrace a combination of both AI and simulation technologies in order to make novel design and production more cost-effective and efficient.
When considering the future of manufacturing and production, it strikes me that manufacturing communities are now becoming incredibly conscious of the all-encompassing change that AI will bring to UK productivity, from its ability to automate routine, non-value added monitoring tasks, removing the need for human intervention, to the decisions it can make about streamlining the design process. The manufacturers we work with on a daily basis are progressively moving to adopt these advances in digital capability and adapting their processes, which is a positive step forwards. However, to harness the real benefits of automation, manufacturers must embrace simulation with the same zeal, as this will be a key enabler in the development of future AI based solutions in manufacturing.
At the Centre for Modelling & Simulation, we are uniquely positioned to be able to offer high value manufacturers a combination of both AI and simulation, which are necessary bedfellows in new applications devoid of real world data. By looking at the way in which the two technologies interact, we can establish why this will save crucial time and money. On a very basic level, simulation enables us to better understand real world problems safely and efficiently by providing a strong dataset from which an engineer can make an informed decision. This alone can diminish months of physical testing of components into a few seconds. Yet, when we bring AI systems into the equation – which when trained, can potentially make those calculated decisions for engineers itself – we can save money on labour and cut time even further. Embracing a balance of these two technologies and understanding the key opportunities they offer will be vital to the future of engineering design and manufacturing.
From working with our partners in the manufacturing space, one previous project springs to mind which really showcased the value and necessity of using AI and simulation together. This collaborative project with the National Composites Centre showed how manufacturers can gather data on new manufacturing methods through simulation to make quick and agile adaptions to the process design. We know that older manufacturing industries have long-standing data-gathering processes, generated by methods such as Statistical Process Control (SPC) from machining. This provides engineers with years of empirical data to inform strategic decisions about the production process and act as the critical source of information to train AI systems. But what about newer industries such as composites or additive manufacturing? How can manufacturers exploit the potential of AI without a backlog of reliable data? By utilising our simulation capability, we were able to run extensive virtual tests to gather large preliminary datasets to provide initial training to the AI system and, based on our improved insight of the process, we were able to inform our partners of the parameters most likely to affect performance or quality before extensive physical trials were initiated.
The autonomous vehicle industry presents an alternative opportunity of how AI and simulation will interact to bring a high value product to market quickly and efficiently. As in most industries, safety is paramount and rigorous requirements are established to provide a challenging and necessary indicator of performance. From working alongside some of the biggest automotive manufacturers in the UK, we understand that the need to carry out thousands of tests and scenarios on a specification of a vehicle is one of the biggest hurdles faced by the developers of such advanced products. These specifications stipulate that an autonomous system would have to travel the order of 11 million miles with only 1 fatality for it to be classified as safe before it can come to market. It is impossible for manufacturers to test out the range of scenarios required physically, as this would take up too much time and be incredibly costly. Instead the AI systems needed to make decisions and control the actions of the vehicle can work with simulation data to carry out comprehensive testing and cut down the product’s time to market.
CFMS enables high value manufacturers to embrace the benefits of working with AI and simulation, bringing both services together to help manufacturers cut costs and drive efficiency. As experts in the field and representatives of the digital engineering community, we understand that the value that this combined technology will bring to UK manufacturers and how it will transform the design process is undeniable. AI and simulation will be the catalyst for a positive transition into a more digitally-enabled and efficient manufacturing industry.