'Artificial intelligence can propel Italian SMEs'

Collaborations with technology partners such as Nvidia and Microsoft to develop software, digital twin and new solutions such as the industrial metaverse are crucial.

Can you give concrete examples?

Our Industrial Copilot tool is already used by many Italian companies for machine design and on production lines. From aerospace to packaging, from paper to food and beverage. Danieli, a leading manufacturer of machines for the steel and metallurgy industries, has benefited greatly from the training of newly recruited engineers who gain the company's historical experience.

What are the advantages?  

The main benefit is the reduction in time to market. Machines and processes are developed much faster. Accuracy is very high: there are no more amnesias or design hallucinations. The level is at least as high as that of the human being. Debugging done by artificial intelligence is much faster and more accurate than in the past.

How does the digital twin work?

The digital twin has become fundamental in the industrial design phase. Let me give another example. Luna Rossa built and tested the entire boat of the last America's Cup in digital. The boat for the new edition is also being designed with our digital twin. So Luna Rossa respected the time constraints for building the real boat and was able to test and improve the mechanics, strength and aerodynamics digitally. Data that it will use when the boat is built.

Can it also be used for existing products or processes?

Yes. Ducati, another example, uses our digital twin to test tyres on the asphalts of the different MotoGP circuits. With the information collected, the compounds to be used for individual races are chosen.

They are all big companies. Are small and medium-sized ones excluded? 

Let's dispel a cliché. AI is not just a tool for big business. Far from it. It is ideal for small and medium-sized companies. It is inexpensive and cuts down on resources that used to be used on repetitive, low-value-added activities. Those funds can be invested in innovation. For example, Gai Bottlers implemented solutions with the industrial metaverse to reduce costs, increase flexibility and service to end customers. Thus it improved competitiveness. But there are dozens of other cases.

Which ones? 

Epf, a robot manufacturer, used the metaverse to train its machines and make them more efficient. Dema, a manufacturer of engine components, used AI to simplify certain assembly processes. The digital and real worlds are more and more closely connected, the tools are within everyone's reach. As never before. Simulation facilitates data collection and data collection is functional for the production or modification of the final product. Artificial intelligence materially enters the physical world. The transfer of information between one world, the digital world, and the other, the real world, the product, is immediate, unmediated and bi-directional. Even information acquired in the field is transferred to the digital twin for reworking.

What can be done to disseminate and govern AI in the industrial system in Italia?

For the dissemination of new technologies we work with the Competence centres, Made and Birex. We must work better with districts and supply chains. In Emilia Romagna, for example, there are all the conditions to amalgamate skills and professionalism and to involve local institutions. We must work more with districts and supply chains, the strong point of our system.

For skills? 

For professional skills, we are trying to transfer resources within the company as well as technologies. People have to be trained and placed in companies. Every time we do a project, we also try to find the resources to run it. Management is very receptive. Paradoxically, we see more resistance in the technical level.

What are the risks for the system?

The risk is too many regulations. Better precise guidelines than a jungle of often overlapping regulations. Consumer regulations cannot be transferred to the production system. AI Act and Data Act are transversal and often contradictory. The government must support those who invest in digitisation and new technologies and help schools that retrain employees. Stimulate use cases, more than big projects. In factories, in buildings. A European model, without chasing the American or Chinese model of mammoth projects. Action is also needed to push the EU to demand equal treatment with non-EU competitors.

Could energy consumption become a problem?

We have to think about the factories of tomorrow. Think about autonomous buildings thanks to renewable energy. AI will optimise energy production and consumption. There is no digitisation without energy but neither is there energy without digitisation. By 2030, to speak of us, all our buildings and factories will be Net Zero.