GENERAL NEWS / 08-01-2024
“Artificial intelligence has the incredible potential to transform the metalworking industry”
Andrés Portela has a very clear perception of the leading role that artificial intelligence has in the development of the metalworking industry, but he also warns of the conditions required for its implementation to be a success. As Aratubo's R&D+i Director explains, it is true that the technological infrastructure is within the reach of any company via the cloud, but before resorting to that, it is essential that the company identifies the challenge they want resolve and find a partner with extensive knowledge of technology 4.0. Otherwise, he warns, “the experience will be in vain”.
Question. The STEEL TECH fair at the BEC recently brought together some of the leaders of the metalworking sector with regard to matters such as technology and sustainability. What is the situation of the sector like in these aspects?
Answer. I would separate the two concepts. On a technological level I would highlight digitisation and industry 4.0, in other words, how the metalworking industry has adapted and implemented technology to improve efficiency, quality and production. To do this, it has implemented concepts such as sensor systems, data analysis, 5G, computer vision, etc. and at the same time there is innovation in processes, in other words, how this technology contributes to developing new ways to explore more efficient and environmentally-friendly processes, such as the reduction of iron ore, the use of electric arc furnace technologies, green steel, low emissions, etc. In addition, there's one thing that still isn't spoken about much, which is related to efficiency in processes and simulations with digital twins, for example. With these technologies we can avoid quality errors and wastage, and make simulations of processes and materials before they go to factory production. Thus the whole process can be optimised from a sustainability point of view; avoiding more wastage and duplicity; saving energy and even staff, minimising carbon emissions, and much more.
Q. And what is the situation in the metalworking sector? Has it already made this technological leap?
A. I think it has. This is all being applied quite significantly because the metalworking sector is one of the big energy consumers. And at the level of energy consumption, AI and industry 4.0 can improve efficiency in all processes. The impact is very powerful.
Q. You speak of industry 4.0, but has industry 5.0 come into play in the sector yet?
A. Industry 5.0 is a natural extension of 4.0 in which we place people at the centre of technology, alongside two other mainstays: sustainability and resilience. Industry 4.0 is based on digitisation, connectivity and process automation, and it's true that the majority of us are still at this stage. Industry 5.0 seeks greater collaboration between people and production systems. In the metalworking sector the industry 4.0 systems such as sensor networks, data analysis, AI and cyberphysical systems are priority tools when seeking more efficient processes. But the big corporations and centres have already made the leap to the next stage. Industry 4.0 is still evolving because one of its enabling technologies, 5G, is still not fully implemented despite a lot of talk about it. In recent months I've participated in various technology forums…, and the conclusion is that 5G is still not in place for private industrial networks.
“Industry 4.0 is not yet completed because one of its enabling technologies, 5G, is still not fully implemented”
Q. But it's is a basic enabling technology…
A. That's true. On a marketing level everybody talks about it but there are very few… actually in our sector we have only been able to verify two or three cases of companies with pure 5G. Today, this implementation basically consists of installing antennas that provide coverage for the whole company because with the current deployment it's not possible. This is in addition to a very significant challenge, cybersecurity: depending on 5G deployed by one single supplier, whichever it is, is very risky. We have to lean towards extending our own 5G network with its own antennas, with its own coverage, and that requires significant investment.
“We have to lean towards extending our own 5G network with its own antennas, with its own coverage, and that requires significant investment”
Q. What potential does artificial intelligence have in the metalworking industry?
A. AI has the incredible potential to transform the metalworking industry. It provides huge, very advanced data analysis capacities; we're talking about big data, analysis of enormous amounts of data, and this is going to impact automation and decision-making in real time. Its greatest potential and impact are in various areas: process automation, especially in predictive maintenance; supply chain optimisation; product quality control with automated visual inspections by computer vision and in production parameter control in real time… It also has a tremendous influence on energy efficiency and sustainability: we must remember that our sector is a big energy consumer and the economic impact it has on business costs is enormous, especially today with the variations in energy prices. Therefore it's essential for us.
“AI has a tremendous influence on energy efficiency and sustainability, and that's essential for us, as we're big energy consumers”
Q. In other words, it particularly affects production costs…
A. Artificial intelligence goes beyond that: it directly affects ongoing improvement, quality improvement, continuous learning through neuronal networks and machine learning… It provides a dynamic of improvements that help us make our processes more efficient and reduce our costs. And to do it we need to develop algorithms that are more and more sophisticated and able to carry out that continuous learning and integrating the new ground-breaking technologies. To sum up, artificial intelligence has enormous potential in the metalworking industry and there is space for continuous improvement as technology evolves and adapts to the specific needs of the sector.
Q. Within the metalworking sector, to what extent are the various key players implementing this technology?
A. There are various scenarios and speeds. The big companies already have these measures fairly well integrated while SMEs go more gradually, and at different speeds.
Q. How does Aratubo apply artificial intelligence?
A. At Aratubo we're a little more advanced and in late 2022 we joined BAIC, the Basque Artificial Intelligence Center. We actively participate in two groups of skills and talent, in which we deal with the roles, aptitudes and training of the AI, and work actively in the projects group, in which we identify different case studies, technological partners and strategic projects to develop collaboratively. At Aratubo we're integrating and implementing simulations into our day to day life, in both product development and manufacture. We have a working group that identifies all the challenges related to manufacture and new materials that we could resolve with technology 4.0. When these challenges have been identified, we prioritise them for scalable development, to integrate them and to adapt the technology in terms of both hardware and internal software. At the same time, we create working groups to integrate people into the new technologies. From there we take on different projects.
Q. Can you mention a few?
A. We have projects under way to improve the logistics and traceability of raw materials and products, to automate quality control and forecast and optimise energy consumption.
Q. Do you think Aratubo is at the forefront of artificial intelligence implementation compared to the rest of the metalworking sector?
A. Yes, we're quite advanced. We're one of the few companies in the steel processing sector that is as aware and mature on a technological level.
“Aratubo is one of the few companies in the steel processing sector that is as aware and mature on a technological level”
Q. Despite these advantages, there are also critics who speak of the risks this tool entails. Is this perception true?
A. Every time there has been a ground-breaking transformation the same doubts come up. It happened in the industrial revolution, moving from steam to electricity, with automation, with robotisation, with computers, and now it's happening with artificial intelligence. It's true that certain types of jobs will disappear, mostly repetitive and with little added value, but new jobs are also created. With these technologies we're going to be able to stop people doing repetitive tasks that lack value, and we'll have the option of offering these people retraining programmes so they can do tasks with greater added value. In addition, a very large proportion of the new jobs created will be technological, which is something very necessary as we have a huge lack of technologists and new talent.
“These technologies will give us the option of offering these people retraining programmes so they can do tasks with greater added value”
Q. And beyond the workforce?
A. Beyond the workforce it's true that artificial intelligence has generated a few ethical risks related to information privacy because automation enables us to work with huge amounts of data that is directly affected by matters of privacy. Nevertheless, these risks are starting to be channelled with new standards. One thing is clear: Artificial intelligence is not ethical or moral, and if the algorithms it uses and the data it processes lack bias the risks will be minimal. We have in our hands the ability and the obligation to regulate all of this to avoid bias when applying this technology. In fact it has already started to be regulated in Europe when the use of pure artificial intelligence was banned in personnel selection processes, to avoid any possible bias.
“AI is not ethical or moral: if the algorithms and the data it processes lack bias the risks will be minimal”
Q. Are we at a turning point, a paradigm change in the industrial sector?
A. I think we are, there's no turning back. Those who don't join this revolution will have no place in the current world nor in the future, but we must take this route with discretion. In my opinion it's vital to take it slowly and not dive straight into the world of artificial intelligence. This technology is here to stay, it has opened up lots of new opportunities, but we have to take care because we can overdose on technology. I'm in favour of progressively incorporating small chunks of technology, scaling them and gaining digital maturity before increasing them. And we should always make this transition alongside our technological partners: we cannot know everything or hoard all the talent, technology and training, but we can count on well-known technological partners that are the best at what they do and can work with us on our projects.
“There's no turning back: Those who don't join this revolution will have no place in the current world nor in the future”
Q. Is technology within the reach of all companies, from SMEs to large corporations?
A. I think it is: the big corporations have the resources that enable them to invest in advanced solutions and even make them drivers of this technology, but SMEs have the possibility of turning to these technological partners. Thus we can have cloud services, platforms that already offer pre-trained algorithms, even projects as a service that allow companies to use the technology only when they need it, instead of purchasing it. And at the level of technological knowledge, companies you can find support in knowledge centres such as universities, technology centres, institutional support lines and technological partners.
“Artificial intelligence is within the reach of all companies, from SMEs to large corporations”
Q. Is it possible to conceive the future of the industrial sector without artificial intelligence?
A. I don't think so. If we always need to go faster, be more efficient, consume more energy and manage more data because machines are increasingly complex, we can't do without that technology that manages such vast amounts of data, that helps us improve business understanding, strategy and planning, and make advanced decisions. With artificial intelligence I can explore the market see where new needs, new challenges and new products are moving and anticipate it in real time. Whoever doesn't do this will be left out of the market.
“It's not possible to conceive the future of the industrial sector without artificial intelligence”
Q. What's needed for a company to get the best out of this digital tool?
A. To work with AI a challenge has to be identified because, otherwise, the experience will be in vain. When the company identifies the challenges, it must check it has the necessary data to tackle it, in terms of quality, amount and availability. Then it has to seek out the talent, both within and outside of the company, in the form of partners or knowledge centres, and check if the technological infrastructure necessary to take it on is available. If it has it, perfect, but if this is not the case, it must find cloud-based services or a technological partner that can offer it. This partner must have a clear understanding of business so that the conclusions drawn are not erroneous; in other words, there needs to be a complete symbiosis between the intelligence of the business, the know-how in the company and the technology. They have to go together.
“They have to identify the challenge, check they have the necessary data, in-house talent or an external partner with a clear understanding of the business, and the necessary technological infrastructure"
Q. This technology also requires the creation of collaborative networks.
A. I think they're essential. For years we've been part of a cluster that is none other than competitors working together to achieve a better understanding of different aspects. Fighting the war on our own is no longer a thing, even for large corporations, because nobody has the absolute truth. We're still in competition in the aspects in which we have to be; But not in technology.
Q. All this work has already gained recognition in Global Innovation Day. What was the use case that you presented?
A. We participated alongside the BAIC in a use case on the prediction of energy consumption depending on hundreds of variables and parameters. In our case, the energy cost is the second largest after the raw material. We have six continuous manufacture lines and the cost of energy shot up. This project has enabled us to know where we are from a point of view of digital immersion in terms of infrastructure, data and the degree of implementation because it's a transverse project. Firstly we needed to know the energy consumption on each level of manufacture depending on the product, with its specific type of steel and special characteristics, based on 120 parameters that we identify. Once we have determined and identified these parameters and patterns we can know in advance the amount of energy we are going to consume every time a different type of material comes in to produce a specific product. When these consumptions change, the second phase comes into play, which consists of crossing it with preventive maintenance, checking which process is failing and even anticipating possible quality faults arising from it all. This project was catalogued as one of the most relevant cases of application of AI in the industry, which we presented in November as part of the Global Innovation Day.
