Agentic AI: Taking New Wave of Industrial AI Seriously

Can you imagine an AI system that, in addition to predicting machine failures, can diagnose the root cause and recommend a solution? And what if, furthermore, you run such a solution autonomously? That is the potential of advanced self-driving systems in the industrial sector, capable of making real-time decisions and solving complex industrial challenges with minimal human supervision, taking operations and business success to new levels. 

Agentic AI combines AI, real-time data, and specialized knowledge. Its objective? To transform the way industries work. This goes hand in hand with one of the recent predictions of the World Economic Forum (1), which states that humans could go from being simple practical operators to becoming AI-enabled orchestrators, in charge of monitoring autonomous systems and providing criteria as required.

New Era, Transformative Impact

Although the origin of artificial intelligence in industry dates to the middle of the last century, when scientists began to investigate the possibility of creating machines that could think and learn like human beings, it is only recently that this technology has begun to represent a real revolution in industrial environments. This is thanks to its contribution in various areas, from manufacturing plants to mobility systems and energy infrastructure, enabling efficient operations even in complex and high-risk environments, and as if that were not enough, meeting industry standards, such as robustness, safety, and reliability.

An example of its use in the industrial sector is predictive maintenance, where it stands as a powerful application that helps to extend the useful life of assets and reduce unplanned downtime in manufacturing plants. It does this by combining AI with data from sensors embedded in machines, such as temperature, vibration, and pressure, to predict equipment failures and, with this information, proactively schedule their maintenance. 

Now, with the new boom in Agentic AI, industrial artificial intelligence is entering a new era. We are talking about systems designed to act with minimal human supervision to make decisions and conduct tasks autonomously, so that they constantly optimize operations to achieve maximum performance. An example is the Siemens Electronics Works Lighthouse factory in Amberg, Germany, which pioneered the use of an autonomous AI agent for quality control. It assists solder paste printing operations, providing them with real-time recommendations on the optimal process parameters, and autonomously adjusting them if necessary, so that the ideal quality of the paste is achieved.

It is no secret that at Siemens we have been leading the industrial AI revolution for years. Representing one of the first Agentic AI systems developed specifically for industrial use, our Industrial Copilots demonstrate what cutting-edge AI is capable of when designed for the real-world challenges of the industrial sector. 

Industrial Copilots are real-time tools that collaborate with the value chain, with their business teams, and participate in one way or another. They can, for example, focus on design, as a creative partner; in planning, they anticipate the future with simulation of scenarios; in engineering, they are writing and debugging automation code, optimizing sequences, taking on repetitive tasks, and much more; in the shop, they are the operator's second brain, providing real-time information that is easy to understand and apply, or taking charge of maintenance, providing personalized service recommendations and helping to keep machines running and customers satisfied.

It is no coincidence that, in the United States and Europe alone, more than 100 companies are already integrating these technologies into their workflow every day: we are no longer talking only about a technological tool that assists and nothing more; now we are talking about intelligence that, integrated directly into the workflow of industries, allows daily decisions to be made smarter, faster, and more well-founded, and all this almost without human intervention.

This technology is not only transforming the way manufacturers program machines, but also the way they turn industrial data into actionable and truthful information, and how they face the challenges they face every day. 

The Next Step

But  innovation does not stop here. Among our latest developments, we can highlight the following:

• Innovations such as the co-development of an Industrial Foundation Model (with Microsoft), which aims to train engineers and operators so that they can solve daily challenges, interacting in their own language with the industrial AI model. At AUDI AG, we also combine AI and industrial computing with vision inspection and virtual workshop automation to create efficient and flexible production lines. To drive these innovations, we foster a powerful partner ecosystem that includes Accenture, NVIDIA, Snowflake, AWS, and more.

• Expanding the collaboration with Microsoft, integrating Siemens Industrial Edge with Microsoft Azure IoT Operations in a seamless connection that unlocks the full potential of AI, analytics, and digital twins thanks to the flow of real-time data from machines on the factory floor to the edge and finally to the cloud. This results in preventive maintenance that minimizes downtime, machines that have optimal performance and higher product quality.

• Acquisition of Dotmatics, focused on helping the pharmaceutical industry deliver smarter, faster, and patient-centric healthcare by creating a connected data workflow at all stages, from R&D to full-scale production. The intention is for companies in this sector to become increasingly digital, leveraging AI, automation, and digital twins to streamline formulation development, optimize formulation production, drive innovation, and accelerate the commercialization of new medicines.

At Siemens, we remain committed to amplifying the transformative power of industrial AI, ensuring that the sector not only keeps pace with innovation but also continues to thrive and expand.

Sources: 

1 - "Frontier Technologies in Industrial Operations: the rise of Artificial Intelligence Agents", WEF.