Reliable workers: Robots go about their business at Loughborough University
BMW and Rolls-Royce are doing it. Airbus has to do it. Germany is spending vast amounts of money on it, under its ‘Industry 4.0’ programme. And a growing number of smaller manufacturing companies in Britain are also now building a business case for automating manufacturing, using robots and other system components to remove labour from the process of making complex assemblies.
This work is crucial to the future of a competitive manufacturing economy, says Professor Mike Jackson, director of the EPSRC Centre for Innovative Manufacturing in Intelligent Automation at Loughborough University. Machines are more accurate and powerful than humans, they maintain quality control over time and they do not get sick (or, rather, “they need less downtime than humans”).
The EPSRC Centre for Innovative Manufacturing in Intelligent Automation was established in 2011 to investigate applications in industry. Loughborough partners with Cranfield University, the Manufacturing Technology Centre, and several companies to corral the UK’s expertise in electronic engineering, machine systems and human/machine factors. This research aims to understand the level of human skill applied to difficult industrial tasks and brings together a unique blend of human factors and mechatronics, which combines electronic with mechanical engineering. It also studies when automation is financially viable, a big issue for smaller firms.
The centre promises to have a big impact on industry, being driven by a definable – and pressing – business need to speed up high-value asset manufacture without losing quality.
Airbus, Rolls-Royce and Controls and Data Services are the industrial partners. That’s no surprise, given the pressure these companies are under to ramp up the production rates of civil aircraft. Airbus’s order book numbers more than 6,000 aircraft.
Robots have transformed productivity in the automotive sector, and their application is one of the main reasons for the relatively recent resurgent fortunes of Britain’s vehicle manufacturing industry – the UK is now a net exporter of cars. The aircraft companies want to emulate parts of this. The EPSRC centre is there to expedite research so that these companies can implement new practices quickly.
The centre is also reaching out to smaller companies. “Through our EPSRC-funded outreach programme, we are building a wider industrial base and provide some benefit to small firms that would not normally get access to this sort of research,” says Jackson.
Musical instrument maker Percussion Plus is an example of a smaller company that has started benefiting from the centre’s work. Managing director Paul Cobbett says: “Being involved with the robotics team at Loughborough and Cranfield universities has been an enlightening experience, turning a seemingly impossible task, of winding our percussion beaters, into what would appear to be a practical possibility. The team have been informative throughout the project. They are an amazing, eclectic group of people.”
The centre studies a wide range of applications of automation, from fixtureless sheet-metal formation to capturing human skill for transfer of tasks to automated systems. A prime example of its work is in adaptive welding. The TIG welding cell at the centre in Loughborough combines technical measurement capability and human factors assessment to map human skill in a very holistic way.
Reviews of manual welding were conducted to inform the design of an intelligent automated welding system that will track the weld gap, monitor performance and adjust the welding arm in response to this feedback, just as a human would. There are two PhD students researching this field.
Untapped potential: The aerospace industry is keen to invest in more automation to speed up its rate of output
The centre’s current study also involves the three-dimensional sensing of the gap between two thin-walled, complex-shaped components that are to be welded together. This includes the selection of sensing strategies, post-weld quality assessment, algorithms for 3D edge detection, welding process monitoring and software components for post-weld analysis.
The recurring theme is to finesse human-robot interactions. “We look at the human skill in a task, then we assess the degree of difficulty in automating it,” says Jackson. “We then devise an intelligent solution where there is likely to be a person doing tasks that a human can perform better than a robot, and a robot doing tasks a human is not especially good at.”
Embedded in this system is autonomy, he says. “The robot can adapt by measuring things in the environment and, with the appropriate computer code written by our system designers, it can react, often in anticipation of expected changes.”
Experts believe that more and better automated manufacturing will create jobs, not lose them. Professor Phil Webb, EPSRC centre deputy director at Cranfield University, says: “You will always have people in the loop – it is a question of how you maximise them.” Key to the centre’s research is how humans work with machines, and the consequences of these interactions on productivity, safety and human welfare. Webb says: “When people work with automation you hit regulatory issues like safety. We have a safety engineer on the team, as well as psychologists and mechatronics engineers.”
With one of its collaborators, the EPSRC centre is looking at large-scale assembly tasks. The research is not relevant to fully automated processes, such as high-volume food manufacture, but in the combination areas. “Where people and robots mix together you get the best of both worlds – the skill and adaptability of humans, and the brute force and accuracy of industrial robots,” says Webb.
The centre is also a great example of how EPSRC-funded research works with the UK’s technology-enabler vehicle, the Catapult, in this case the Manufacturing Technology Centre (MTC), which helps companies get to commercial viability from a Technology Readiness Level (TRL) 3 or 4 prototype.
Working with Professor Ken Young at the MTC, it became clear there were several fundamental automation issues to solve at low TRL levels. These informed the EPSRC centre on what to research. The results were fed back to the MTC and then to industry, to be used in production. An example of a low TRL problem that the researchers tackle is effective risk assessment protocols. Webb says: “There are no fully defined or accepted risk assessment methodologies that can deal with large-scale collaborative automation. Current risk assessments will say ‘this is dangerous, we will mitigate the risk by putting a cage around it’. We can’t have cages, so we need other ways to deal with risk.”
The team analyses failure rates of the systems, and then what happens when you have an unpredictable person in the process. This illustrates the complex ergonomic conundrums that the centre is tackling, to design more productive, quicker and safer manufacturing systems to keep the UK at the head of the race.
Jackson says the essential work of the centre is getting through to government because of lobbying by the British Automation and Robot Association, which helped to secure the Automating Britain fund from the Department for Business, Innovation and Skills in 2012. But he is concerned that a new government might switch off funding for intelligent automation through lack of understanding of its significance.
“Governments must try harder to understand the need for using this technology and building machines,” says Jackson. “Automation is not going to slow down.”