Researchers Luigia Carlucci Aiello and Marta Lagomarsino on Rai News
Ahead of the 19th edition of RomeCup (28–30 April 2026), our media partnership with Rai News brings us a valuable intergenerational dialogue between two leading figures in Italian research: Luigia Carlucci Aiello, ambassador for Sapienza University and considered the ‘mother’ of artificial intelligence in Italy, and Marta Lagomarsino, a postdoctoral researcher at the Italian Institute of Technology and winner of the first edition of the Most Promising Researcher in Robotics and Artificial Intelligence award, with a project on bio-cooperative robotics aimed at improving the quality and safety of working life. Two generations, two research paths, one fundamental question: how can we build technologies capable of remaining at the service of people?
Luigia Carlucci Aiello begins with one of the most sensitive issues in artificial intelligence: the supposed neutrality of algorithms. Technology, she points out, never emerges from an abstract space. It is designed by people, fed by data collected and organised by people, and carries with it the knowledge, limitations and biases of those who build it. “Both the algorithm itself and the collection and preparation of the data on which the algorithm runs are heavily influenced by the culture, knowledge and, consequently, the ignorance of the person designing it,” explains Carlucci Aiello. “This is then reflected in what are now known as biases, that is, prejudices and preconceptions that the software carries with it.” The issue is not merely technical, but cultural. Artificial intelligence systems can produce incorrect or questionable responses, which may not be acceptable from the recipient’s perspective, precisely because they incorporate the viewpoint of those who designed or trained them. Sometimes the error can raise a smile, as in the case recounted by the professor: an automated system, whilst translating her CV into English, changed some research experiences to the masculine form, probably because they were historically associated with predominantly male careers. But in other contexts, she warns, errors can be far more dangerous, especially when the user fails to recognise them.
Hence the value of the human in the loop, the principle that humans must be able to understand, control and, when necessary, interrupt a system’s behaviour. “The more autonomy the system has,” emphasises Carlucci Aiello, “the more important it becomes that, at a certain point, a human can take back control and impose a certain behaviour, blocking this autonomy which might otherwise become dangerous”.
The distinction between automatism and autonomy therefore becomes crucial. A lift responds to a deterministic command; an autonomous system makes choices whose outcome may be difficult, if not impossible, to predict. For this reason, the possibility of human intervention is closely linked to understanding what the software is doing.
In her conversation with RaiNews.it, Carlucci Aiello also extends an invitation to the younger generations, and in particular to young women choosing to pursue research in artificial intelligence and robotics. “The best way to predict the future is to invent it”, she recalls, quoting Alan Kay. She adds: “If you want to help create a future in which technology also incorporates a female perspective, now is the time to saddle up and set off on this adventure”.
It is an invitation to enter the places where the future is designed, without fear of complexity or failure.
“When you tackle a research problem, if you don’t fail 90% of the time, it means it’s not really research”, she adds. From this perspective, failure is not an obstacle but part of the method: you learn, you try again, you keep aiming high.
The second part of the interview is dedicated to the research of Marta Lagomarsino, who works on developing artificial intelligence algorithms to detect the user’s psychophysical state, preferences and skills by observing movement patterns. The body thus becomes a valuable source of information for designing robots better able to adapt to people.
“Movement constitutes a highly significant channel of information,” explains Lagomarsino, “which artificial intelligence algorithms can analyse and process in a very effective and useful way.” When artificial intelligence is then equipped with a body through robotics, the concept of physical AI takes shape: a technology capable not only of interpreting data, but also of acting in the physical space to generate “a targeted benefit for people”.
The aim of the research is to enable robots to learn new skills and adapt their behaviour to reduce the workload, meet users’ needs and create “a more effective and fluid collaboration between humans and robots”. This is a particularly important prospect in occupational contexts, where an ageing workforce, a shortage of skilled staff, musculoskeletal disorders, stress and burnout make it increasingly urgent to design technologies capable of supporting people’s well-being.
Lagomarsino emphasises a key point: for these technologies to be truly useful, they must become more adaptive, intuitive and accessible even to non-expert users. Body language analysis allows information to be gathered continuously and non-invasively, by observing spontaneous movements during activities. On this basis, machine learning algorithms and optimisation techniques can help robots adjust their pace and operating modes, with the aim of increasing productivity “without compromising the user’s well-being”.
The outlook for the future is already very concrete. “The first application I see on the horizon is the use of these robots to assist workers in companies,” says Lagomarsino. “We would like these systems to become more intuitive, accessible and attuned to humans in the near future, so that they can work side by side with them.”
Alongside the industrial sector, the researcher is particularly interested in health and inclusion. Artificial intelligence and robotics technologies can support people with mobility impairments in their daily lives and in their pathways to employment or return to work. “Body language analysis is a rich source of information,” she notes, “because it allows us to identify abnormalities in movement, signs of deterioration linked to ageing, or progress in functional recovery. Intelligent decision-making systems can therefore dynamically adapt the level and mode of support according to individual needs.”
The exchange between Luigia Carlucci Aiello and Marta Lagomarsino captures the deeper meaning of RomeCup 2026: bringing together generations, skills and visions to explore the future of innovation. Not just what technology can do, but how it can be designed, governed and directed towards people’s well-being.
Organised by the Fondazione Mondo Digitale, RomeCup brings together schools, universities, research centres, companies and institutions every year to explore the frontiers of innovation. The 2026 edition is expected to attract over 4,000 participants, featuring talks, robotics competitions, demonstration areas, workshops and the fourth edition of the Most Promising Researcher in Robotics and AI award.
From 28 to 30 April, RomeCup thus returns as a laboratory of the future: a space where human and artificial intelligence, research and education, experience and young talent come together to build technology that is more inclusive, responsible and closely aligned with people’s lives.
