Industrial AI Is the True Value Driver of the AI Revolution, Says Siemens Exec

Industrial applications of artificial intelligence (AI) are set to deliver the most significant value in the technology’s evolution, according to Dr. Xiao Song, Global Executive Vice President of Siemens and CEO of Siemens China.

Speaking at TMTPost’s “Talk to the World Forums” during the CES 2026 on Wednesday, Xiao said consumer-facing AI, such as chatbots and interactive tools, generates traffic and visibility but cannot match the reliability and impact of industrial AI.

“At present, the role of AI is evolving as applications are being implemented,” Xiao said. “It is shifting toward real-world system scenarios, developing into intelligent collaborative entities that focus on the ‘last mile’ of application. Industrial AI, by contrast, is a tougher nut to crack—it requires foundational skills, domain knowledge, and extremely high reliability.”

Xiao’s comments reflect a broader industry trend emphasizing the integration of AI into industrial processes, where it can optimize production, reduce costs, and improve efficiency.

The potential of industrial AI was highlighted at CES 2026 in Las Vegas, where NVIDIA founder Jensen Huang introduced the concept of “physical AI”—embedding AI into real-world industrial systems. Siemens, recognized by NVIDIA as a key partner, showcased its Digital Twin Composer platform, which integrates simulation, real-time engineering data, and digital twin technologies to support industrial operations.

“Digital Twin Composer allows the production of any product to be mirrored in a virtual world,” Xiao said. “During production and design, virtual and real-world products communicate in real time. High-quality data generated in the virtual world is continuously accumulated, improving efficiency, quality, and productivity. PepsiCo has used the technology to simulate upgrades at U.S. plants and plans global deployment.”

Xiao compared industrial AI’s potential impact to that of electricity, noting Siemens’ 170 years of industry expertise and high-quality data. “Industrial AI cannot be rushed. It requires careful collaboration, precise application, and ecosystem-wide efforts,” he said.

He stressed the difference between industrial and consumer AI. “Consumer AI emphasizes user experience, whereas in industrial settings, errors can halt production or compromise products. Reliability is paramount. Every vertical industry has unique parameters, so industrial AI must be precise, domain-aware, and application-focused,” he said.

Siemens is leveraging consumer AI innovations while tailoring them for industrial use. Startups are applying Transformer architectures to generate automotive test cases, a task that previously required large engineering teams. Xiao said such collaboration demonstrates how AI can improve efficiency without compromising standards.

Although it was his first CES attendance, Xiao has followed the event for over 20 years due to his automotive background. He noted the show itself is shifting focus from consumer electronics to industrial intelligence and digitalization. “CES is increasingly focused on how AI can be embedded into both hardware and software to drive industrial value,” he said.

Siemens’ CES 2026 presentation highlighted the integration of AI, digital twins, and physical systems. Xiao said the Digital Twin Composer enables real-time interaction between virtual and real production lines, allowing issues to be identified and addressed before physical implementation.

Xiao also discussed the concept of the “industrial brain”—AI systems that oversee and optimize complex industrial processes—and the emerging field of embodied intelligence, which combines AI cognition with robotics. Digital twins can simulate scenarios and accelerate the adoption of robotics in industrial applications, he said.

A key element of Siemens’ strategy is ecosystem collaboration. The company plans to invest 1 billion euros over three years to expand its industrial AI ecosystem. “We aim to establish an industrial foundation model that will enable partners to develop Industry AI Agents. We engage with clients of all sizes, providing or co-creating solutions to address specific pain points,” Xiao said.

Through its Siemens Xcelerator platform, the company integrates its technologies with co-created solutions from clients and partners, enabling faster AI adoption among small and medium-sized enterprises (SMEs) in China.

Addressing concerns about AI replacing human workers, Xiao said the technology complements human expertise. “AI will not replace humans, but those who master AI will replace those who do not. Experienced workers are more valuable than ever, as AI can codify expertise, preserve knowledge, and optimize operations,” he said, citing an example with China Fifteenth Metallurgical Construction Group, where AI algorithms based on veteran experts’ experience improved copper smelting processes.

Xiao said the first measurable benefits of industrial AI are improvements in quality, efficiency, and cost reduction. “Industrial AI is not theoretical—it delivers value. It may take three years to achieve broader breakthroughs, but the market potential is enormous,” he said.

China’s industrial landscape, with its breadth, depth, and proactive adoption of new technologies, provides fertile ground for AI applications, Xiao added. “With government initiatives to open up more industrial scenarios, China is well-positioned to lead in AI adoption,” he said.

Despite industrial AI’s potential, Xiao cautioned against rushing implementation. He said building a comprehensive “industrial brain” in one step is risky. Siemens is instead focusing on foundational models and targeted applications to enable gradual AI deployment across sectors.

Xiao compared industrial AI adoption to autonomous driving, estimating the sector is currently at an L1 stage, with significant room for growth. “We are only beginning to explore what’s possible. The focus should be on converting uncertainties into certainties, which is the only way to create real industrial value,” he said.

Siemens’ combination of industrial expertise, advanced digital twin technologies, and ecosystem collaboration positions the company at the forefront of the industrial AI transformation. Xiao said the convergence of virtual and real-world processes and the deployment of AI in industrial systems represents a new era for industry.

Siemens, founded over 170 years ago, has invested heavily in digitalization and AI-driven technologies, including predictive maintenance, intelligent production systems, and digital twins. Its partnership with NVIDIA supports the integration of AI and digital twin technologies across multiple industries. Through the Siemens Xcelerator platform, clients and partners can co-create solutions, deploy AI Agents, and access foundational models for industrial applications.