New-Tech Europe | Q3 2026 | Digital Edition

"

modern manufacturing. Artificial intelligence adds a new capability: the ability to interpret relationships between previously disconnected sources of information and transform them into operational knowledge. For European manufacturers facing rising energy costs, labour shortages, increasingly complex supply chains and new regulatory requirements, this capability is rapidly becoming a strategic advantage rather than simply another software upgrade. The technologies enabling this transformation are already emerging. While each addresses a different challenge, together they form the foundation of the next generation of intelligent manufacturing. The Five Technologies Transforming Manufacturing: Foundation Models: Creating a Common Industrial Knowledge Layer For years, industrial AI applications have been designed to perform specific tasks. One model detects surface defects, another predicts equipment failures, while a third optimises production schedules. Each application performs well within its own domain but has little understanding of the broader manufacturing environment. Foundation Models introduce a different approach. Rather than developing a separate AI model for every industrial application, Foundation Models establish a shared knowledge layer capable of understanding multiple forms of engineering information simultaneously. Manufacturing knowledge exists in many formats. Sensor measurements, CAD drawings, technical documentation, PLC programs, maintenance manuals, quality reports, simulation results and ERP data all contribute to operational decision making. Traditionally, these information sources have remained isolated within specialised software platforms. Foundation Models make it possible to connect them. Instead of analysing a vibration alert as an isolated event, an AI system can combine maintenance history, production schedules, engineering documentation and equipment specifications to provide engineers with a much broader understanding of the problem. The objective is not to replace engineering expertise, but to make it significantly easier to access and apply. As multimodal AI continues to mature, Foundation Models are expected to become the common knowledge layer supporting virtually every industrial application. AI Agents: From Information Retrieval to Workflow Coordination If Foundation Models provide industrial knowledge, AI Agents put that knowledge to work.

In the near term, AI Agents are unlikely to replace engineers. Instead, they will reduce the effort required to analyse increasingly complex manufacturing environments, allowing specialists to focus on decision making rather than information gathering.

"

Manufacturing decisions rarely depend on a single dataset. Diagnosing a quality issue may require maintenance records, inspection reports, supplier documentation, engineering changes and production history. Gathering this information often consumes more time than solving the problem itself. AI Agents are designed to coordinate this process. Rather than answering isolated questions, they pursue operational objectives by collecting information from multiple systems, selecting appropriate tools and presenting engineers with structured recommendations supported by relevant evidence. The emphasis shifts from automation toward intelligent coordination. In the near term, AI Agents are unlikely to replace engineers. Instead, they will reduce the effort required to analyse increasingly complex manufacturing environments, allowing specialists to focus on decision making rather than information gathering. This collaborative model reflects the priorities of industrial manufacturing, where transparency, traceability and engineering oversight remain essential requirements. Digital Twins: From Simulation to Operational Intelligence Engineering has always relied on models. Digital Twins extend that concept by creating continuously updated virtual representations of physical assets, production lines and manufacturing facilities. Unlike conventional simulations, Digital Twins evolve alongside their physical counterparts, receiving real time information from sensors, industrial controllers and enterprise systems. This allows manufacturers to evaluate operational changes before implementing them on the factory floor. Engineers can test production scenarios, optimise energy consumption, evaluate maintenance strategies and validate process improvements without interrupting ongoing operations. The value of Digital Twins becomes even greater when combined with artificial intelligence.

New-Tech Magazine Europe l 21