Portugal develops AI ‘digital twins’ to predict disasters and improve public services

Portugal is developing advanced “digital twin” technology powered by artificial intelligence to help predict natural disasters, improve urban mobility and streamline public services, according to the Minister for the Presidency and State Reform, Gonçalo Matias.

Speaking in Lisbon following a meeting of the State Simplification and Technology Network, the minister said the government is exploring the use of AI-driven digital replicas of real-world systems to support decision-making across a range of sectors.

The technology is being developed by the government’s Agency for the Technological Reform of the State (ARTE) without relying on foreign companies. It is a move aimed at strengthening Portugal’s technological sovereignty, explained Matias.

As for the ‘digital twin’, this is a virtual model that mirrors a real-world object, system or process. By integrating real-time data from sensors and connected devices, the technology can simulate scenarios, identify risks, predict failures and test solutions before they are implemented in reality.

Matias said potential applications include disaster forecasting, and urban mobility management, allowing authorities to anticipate problems and respond more effectively.

The minister also confirmed that Amália, Portugal’s first sovereign large language AI model, is now complete and will be officially unveiled next month.

“Amália, the Portuguese-language large language model, will be ready and presented in July and will therefore also be a tool for the use of artificial intelligence within public administration,” he said.

Amália has been designed to handle sensitive information and support strategic sectors ‘where data sovereignty is critical’.

“It is a sovereign model, which allows us to address very specific use cases, whether in healthcare, the Armed Forces, or other areas involving sensitive information,” said the minister, adding that the government is currently assessing AI applications across ministries to improve resource management, accelerate licensing procedures, speed up public procurement processes and enhance the quality of public services.

While describing Amália as an important step forward, Gonçalo Matias stressed that the platform is not intended as a universal solution.

“It will not solve every problem, but it is a tool available to us for specific use cases where there is a strong need for sovereignty,” he said.

The meeting also focused on an ambitious AI training programme for Portugal’s public administration. Under the government’s Digital Skills Pact, up to €80 million has been earmarked to improve AI literacy and technical capabilities among public sector workers, with an initial investment of €25 million already planned.

Training will be delivered through a combination of in-person and remote programmes.

Matias also highlighted potential financial benefits from the State Simplification and Technology Network’s work, saying that standardising technology procurement across government departments could generate savings of around €300 million.

The savings are expected to come from ‘harmonising procurement specifications and negotiating unified contracts with major technology suppliers, enabling the Portuguese state to leverage greater purchasing power while reducing duplication and inefficiencies’, he added.

Source: LUSA