Diego Droghini joins AIPIA's Technical-Scientific Committee

Diego Droghini joins AIPIA’s Technical-Scientific Committee with effect from 20 September 2026. A PhD in Electronic Engineering, he is an AI Specialist and Machine Learning Research Engineer, and brings to the Committee more than ten years split between applied research and putting artificial intelligence systems into production in industrial settings.

The Technical-Scientific Committee is one of the four bodies set out in article 6 of the Statute, alongside the President, the Board and the Members’ Assembly. It oversees the methodological validity and scientific coherence of the association’s work, draws up guidelines, ethical standards and professional good practice, and reviews the quality of the projects AIPIA promotes.

Research and production, in the same person

His profile holds together two things that rarely coexist. On one side, peer-reviewed machine learning research: a dozen international publications across journals and conference proceedings, built around computational audio signal processing, and a Best Presentation Award at the 2018 International Conference on Pattern Recognition and Artificial Intelligence. On the other, commissioning on the factory floor, where an error carries a measurable cost and a false alarm carries a different one.

It is a combination that serves a technical body on two distinct fronts. The first is the methodological assessment of projects: telling a solid result from one that only holds on the test data is a craft learned by publishing and being reviewed. The second is drafting guidelines that hold up outside the laboratory too, where conditions are never those of the test bench.

Generative-AI governance inside a company

Since November 2023 Droghini has been AI Specialist at ICS Technologies, where he leads company-wide adoption of generative artificial intelligence: defining governance frameworks, operational usage guidelines, and the criteria by which language models enter business and engineering processes. The stated approach is not normative from above but experimental: identify the processes where impact is measurable, verify in the field, and only then consolidate practice into a guideline.

This is precisely the ground on which the Committee develops standards and professional good practice for the artificial intelligence community. The same role covers the infrastructure connecting AI clients to company systems, with MCP connectors to the ERP, the document system and the simulation tools, and internal training: sessions for all staff on the considered use of generative tools, and specific developer tracks on coding agents.

From audio signal to in-line quality control

Between 2018 and 2023, as AI Research Scientist at MFL X, he devised and developed a real-time quality control system for the production of electrical cable sheathing: defects have to be caught while the line is moving, because checking downstream means scrapping material already made, and a false alarm means stopping the plant for no reason. The method is the subject of patent document US20200408696A1, filed in 2019 in the names of Vesperini and Droghini.

In the same period he designed an IIoT web application for production line monitoring and coordinated cross-disciplinary teams of developers, automation engineers and mechanical engineers: another piece of craft a technical body draws on when different competences have to speak to each other about the same project.

Agents, and their limits

Alongside his work in the company, Droghini carries out independent, continuing research on agentic engineering and agent-assisted programming. The test bed is not an experiment but a real product: Bigfood, which began in 2024 from a food laboratory’s need for traceability and is today used by customers for HACCP compliance and day-to-day management in small and medium food businesses, built entirely with tools of this kind.

Since 2026 he has also worked on Panorama, commissioned by the Institute of Tropical Medicine in Antwerp: a system supporting differential diagnostic reasoning where access to tests and laboratories is limited. The methodological approach is as relevant to the Committee’s work as the result: every relation in the knowledge base carries the source it was extracted from, assessments are proposed by the agent but decided by the medical team, and in operation no decision is taken by a model.

The Technical-Scientific Committee holds no powers of representation: legal representation of the association rests with the President under the Statute, and no member of the Committee can bind AIPIA. The body’s contribution is technical and methodological.

Where to read more

The Diego Droghini profile sets out his professional background in full. The current membership of the body is on the Technical-Scientific Committee page.