What is Physical AI according to DGS?

Physical AI is artificial intelligence that enters the physical world and transforms the relationship between machines, data, processes, and people. In this article, DGS расскаtells the shift from automation to the learning factory, with an in-depth piece by Francesco Di Bianco, Physical AI Offering Leader at DGS, published on Agenda Digitale.

The article examines Physical AI as a convergence point between artificial intelligence, robotics, sensors, and industrial systems. This evolution takes AI beyond content generation and forecasting, toward systems capable of perceiving the environment, understanding context, making decisions, and acting through a continuous feedback and learning loop.

How does the shift from an automated factory to a learning factory work?

The transition described by Francesco Di Bianco concerns systems that no longer simply execute tasks, but learn from the environment and operational interactions. Physical AI makes it possible to produce in a way where robots, data, and processes work together dynamically, with the ability to adapt and learn continuously.

From computer vision to natural language, and even tactile information, multimodal models enable cognitive robots to recognize objects, interpret operational situations, adapt their behavior, and collaborate with operators. Digital twins and simulation environments also make it possible to train these systems on thousands of scenarios, speeding up learning and reducing risks before deployment on the factory floor.

What is the role of MES, ERP, and PLM in Physical AI?

The value of Physical AI becomes fully apparent when its capabilities are integrated with the company’s digital ecosystem. Connection with MES, ERP, and PLM transforms the robot from an autonomous unit into an active node in the production process, able to receive priorities, return operational data, and detect anomalies.

This integration helps continuously optimize production, supply, quality, and design. In this way, Physical AI is not just an isolated technology, but becomes part of the company’s decision-making and operational chain, with direct effects on industrial process management.

What conditions are needed to govern this evolution?

According to Francesco Di Bianco, Physical AI requires specific conditions of technological and organizational governance. Industrial cybersecurity, data quality, protection of IT and OT environments, training, and reskilling become structural components of a roadmap that involves technology, organization, and operational security.

These elements are crucial because the evolution described in the article is not only about adopting new solutions, but also about the company’s ability to manage them securely and coherently. The roadmap outlined by DGS therefore brings together skills, environmental protection, and the quality of the digital infrastructure.

What scenarios does Physical AI open up for businesses?

Physical AI opens the door to factories capable of self-regulation, processes that learn from experience, and information systems that are increasingly connected to decisions made on the shop floor. It is a change that pushes companies beyond the logic of a single use case and toward building an integrated, secure architecture focused on continuous improvement.

The article published on Agenda Digitale presents this transformation as an ongoing shift, in which artificial intelligence enters physical and industrial processes. To read the full insight, the link to the article on Agenda Digitale is available.

Read the full article