In the latest episode of the Future Ready Podcast, host Spencer Acain sat down with Samir Desai to continue their ongoing conversation about how Siemens is approaching industrial artificial intelligence (AI) across its software portfolio. This episode dove deeper on data security and the rise of AI agents. You can check out the podcast through the player below, or keep reading for a summary of the conversation!

Common Hurdles in the Race to Industrial AI

As more companies pursue AI, a common pattern has emerged in the enterprise software market. Vendors frequently approach customers proposing that they export product lifecycle management and operations data into an external data lake, offering to handle the training of AI models and data contextualization going forward. Samir argues that this seemingly convenient approach can actually hamper the success of an AI strategy.

There are two key problems. The first is data freshness. In a modern engineering environment, data is dynamic. Designers are constantly iterating, updating models, and refining specifications. The moment that data is exported to an external system, it becomes untethered from the underlying source of truth and can thus drift from the reality it is meant to represent. By the time an AI model acts on it, the underlying reality may have changed. For companies operating at the pace that modern product development demands, this lag can lead to decisions being made on outdated information.

Secondly, and potentially more critically, this arrangement also introduces data security, control and access issues. Samir explained how industrial customers build layers of robust security and access controls around their systems of truth, such as product lifecycle management systems, that ensure the right data is available at the right time based on the user role and other requirements. External data lakes present significant unknowns, particularly around access permissions.

For AI systems, direct connections to data systems offer better performance and greater security for customers and partners. Siemens’ AI fabric, for instance, connects directly to PLM or other systems, rather than extracting data from them. The direct connection enables AI solutions to adopt the same permissions as the underlying system and ensures that data is always up to date.

Moving to AI Agents

Samir and Spencer then shifted to discuss the rise of AI agents in enterprise applications. These more sophisticated AI systems can respond to sequences of actions, making intermediate decisions and interacting with other systems autonomously to complete complex workflows. Desai explained that Siemens has been actively building toward this capability, with one of its earliest and most significant moves being the development of the Fuse EDA AI system.

Fuse serves as an AI framework specifically designed for Siemens’ Electronic Design Automation portfolio. It provides the building blocks for constructing agents, including critical infrastructure for agentic information discovery, task execution and interoperability. Most organizations have standardized on a mix of tools, including ERP systems and other platforms.

Domain Knowledge Creates Differentiation in Industrial AI

Spencer and Samir rounded out the conversation by diving in to what sets vendors apart in the industrial AI space. In Samir’s words, “Very simple, domain knowledge.”

In this regard, Siemens is unique. Over four decades developing and refining industrial software, Siemens has developed a deep understanding of how customers in roughly a dozen industries use those solutions, the challenges they face and how they characterize success. This context, developed over time, allows Siemens to move faster, target the right problems and deliver AI that is built for purpose.

This is a meaningful distinction in a market where many AI offerings are broad by design. Industrial applications, in contrast, demand precision, reliability, and a deep understanding of the physical and operational realities involved.

Industrial AI Built on Strong Principles, Tech

In sum, Samir and Spencer’s discussion explored how secure and reliable AI solutions depend equally on sound principles and advanced technologies. Secure data, connectivity and the integration of deep domain expertise emerged as core to the development of effective industrial AI.

Check out the Future Ready Podcast to learn more about Industrial AI, the Digital Twin, Connected Automation and more key technologies transforming the future of industrial technology.

---

Siemens Digital Industries Software helps organizations of all sizes digitally transform using software, hardware and services from the Siemens Xcelerator business platform. Siemens’ software and the comprehensive digital twin enable companies to optimize their design, engineering and manufacturing processes to turn today’s ideas into the sustainable products of the future. From chips to entire systems, from product to process, across all industries. Siemens Digital Industries Software – Accelerating transformation.