At the University of Patras campus, seven kilometres from the city centre, the future becomes present every day, in the Artificial Intelligence (AI) product maturation hub, one of only eight created across Europe, in the framework of the Community project AI-MATTERS. The uninitiated visitor may be ‘lost’ in front of a series of robots, devices and machines, whose operation seems incomprehensible, but most of the 100 researchers working in the hub, in its physical and digital infrastructure, are not consumed by obscure theories: they are working on technological solutions, all practical, aimed at enhancing the competitiveness of Greek businesses, especially small and medium-sized enterprises. This innovative hub is the result of the collective effort of the Laboratory of Production Systems and Automation (LMS) of the University of Patras and the Teaching Factory Competence Center (TF-CC), on whose premises it is located.
What stage is the whole project at? As Professor Georgios Chrysolouris, coordinator of the Greek hub, explains to APE-MPA, the hub is now at the stage of providing services, with more than 15 AI products already maturing. How are these products maturing? By testing and experimenting in practice. There are already five robotic cells on site, equipped with industrial robotic arms of large (up to 170 kg) and small (up to 10 kg) load, collaborative robotic arms, mobile robotic platforms, vision and force/torque sensors, accelerometers, thermal cameras, tools for robotic welding, etc. In addition, the two organisations ‘running’ the whole project have equipment for advanced simulations, including those performed in virtual reality environments.
‘Key’ to the operation of the hub is, according to Mr. Chrysolouris, the expertise gained from the LMS’s long-term participation in more than 200 European and national projects, but also the research experience of the Laboratory’s members, who for more than 30 years have been developing solutions in collaboration with industries such as VDL and PSA groups, the Dutch semiconductor machinery manufacturing giant ASML and Philips, as well as Greek companies such as SIDENOR, KLEEMANN, CALPAK.
Which businesses can benefit and how quickly do projects mature?
“The 15 TN products currently maturing in the node include solutions for automated industrial quality control, intelligent robotic platforms for flexibility, predictive maintenance systems, digital twins for production simulation and optimization, and data analysis/processing tools. Some of the technologies are already being evaluated in real industrial production conditions, in collaboration with companies developing AI solutions. Several trials and experiments have already been carried out with the participation of industrial partners, and new solutions are continuously being developed.
Which businesses can benefit? The AI-MATTERS project is aimed at both technology providers and end, industrial users, with a focus on SMEs, but without excluding collaboration with larger industries. Interested companies can get in touch with the hub’s executives and once their request has been evaluated, development and testing will follow and finally a report with the results of the work. “The process is relatively simple so that we can attract as many companies as possible. The duration of the development and testing stage depends on the complexity of the solution and the specific requirements of the industry, but usually ranges from three to six months,” discloses the professor, who shows us around the hub.
The four major challenges in implementing AI in Greek industry and what the hub can offer
The implementation of AI in Greek industry has to face, according to Mr.Chrysolouris, many different challenges. Firstly, high initial investment costs are usually required, especially for AI solutions, which are often not guaranteed to perform as expected, as they are new applications that have not been tested in a sufficiently large number of cases. ‘For this reason, AI-MATTERS enables what we call “test-before-invest”, i.e. the ability to test AI technologies before investing in them to reduce risk,’ he says.
A second challenge is that, in many cases, the data from sensors, production machines, cameras, etc., required to develop AI models based – for example – on Machine Learning, are either not enough in number or of low quality. ‘To solve the problem of data quantity and quality, we enable the creation of synthetic data, using simulation technology as well as specific algorithms,’ explains the professor. Another obstacle is the requirement for skilled personnel who can handle AI applications. To this end, the Teaching Factory Skills Centre has a training and specialisation programme specifically aimed at industry workers, such as machine operators, production engineers, etc. “Finally, AI applications require interfacing with IT systems already installed in industry. This is usually a complex problem because of the many different IT systems, the data they manage and the fact that they “speak” different “languages”. We can and do provide consultancy services to help companies interface their systems with the new AI systems,” he adds.
Why put AI in the factory?
But why is it important to ‘get’ AI into the factory, but also technologies such as 3D printing and virtual and augmented reality (VR/AR)? In very simple terms, for reasons of saving time and money, training, collaboration and environmental protection. For example, AI is already being used to analyse sensor data and predict failures in machinery and equipment, in order to make better use of machines and improve productivity. It can optimise the production process, adjusting parameters in real time to maximise efficiency and reduce waste. In addition, AI powers robots performing complex tasks in industrial environments, improving safety and productivity.
And 3D printing is being used to rapidly and economically prototype products, as well as to manufacture parts that are difficult or impossible to make using traditional methods, improving development time and reducing costs. The combination of AI in the design of 3D printed products can achieve the production of innovative products with highly technical characteristics for the automotive and aerospace industries.
Virtual Reality (VR) technology is used to train workers in dangerous or complex tasks, improving safety and efficiency. For example, Boeing uses VR to train staff for aircraft maintenance tasks. VR can also make an extremely positive contribution to the design and simulation of 3D products, improving communication and collaboration between engineers and designers, as implemented by Siemens. AR is used to provide real-time instructions, data, and information to workers, improving efficiency and productivity.
AI could potentially create inequalities, but at the same time it can have a balancing potential, as it offers useful tools to smaller ‘players’. How could it affect the competitiveness and productivity of Greek industry? “The impression we have formed, from watching Greek industry closely, is that despite the significant investments made in digital transformation and industrial excellence projects, ultimately most and possibly the most important decisions are made based on people’s experience and intuition. Therefore, AI could contribute by helping industry executives to make more effective decisions, based on data and analysed in measurable results. The cost of integrating technology can be a major constraint for firms. However, the “democratisation” of AI solutions can also provide equal access to these technologies for small and medium-sized enterprises (SMEs). And here we come back to the “heart” of the AI-MATTERS programme, which enables SMEs to test and deploy the technology at virtually no cost on their part, as the costs are covered by the programme,” he notes.
The AI-IGNITE competition
Meanwhile, the AI-IGNITE competition is in the pipeline, which aims to support new and start-up companies to develop, implement and evaluate AI solutions for the manufacturing industry. Five start-ups or SMEs will benefit under the competition, and which ones will benefit will be based on the viability prospects and technical feasibility of the AI implementation they propose. AI-IGNITE is organised by the EIT Manufacturing Region South East, in collaboration with the Greek hub of AI-MATTERS. The competition will open on 15 June, the deadline for applications will be 1 August and the finalists will compete at an event to be held in September. The winners of the competition will receive a cash prize and will have access to the Greek node to further refine their solution.
The AI-MATTERS project was recently at the centre of an event on AI in industry, where, in addition to Mr Chrysolouris, speakers included Zoe Arkouli, an associate of the LMS Laboratory, Dr Kosmas Alexopoulos, also a member of the Laboratory, Dr Sotiris Makris, executive director of the competence centre, Dr. Jenny Livadarou, associate of the competence centre, Dr. Niki Koussis, director of EIT Manufacturing South East and Dr. Vivian Sofou, executive of EIT Manufacturing South East.
Alexandra Guta
Are you interested in one of our services? Do you want to know more on how AI-MATTERS works and what we can do for you? Get in touch with us!
