At this week’s Applied AI for Industry Transformation forum, speakers outlined how advances in industrial, agentic and physical AI are reshaping work and production as we know it
At the Applied AI for Industry Transformation forum, speakers from McKinsey & Company, Siemens Taiwan and Advantech said the next phase of AI adoption will be defined not by chatbots, but by real-world industrial applications.
“For the past 50 years, we’ve operated our businesses on a simplifying assumption — that the core engine of productivity is human labor with some help from technology,” McKinsey & Company partner Dr. Denise Lee said. “That era is over.”
In the job market, AI-related skills are the fastest-growing category in demand, affecting roles beyond STEM, she said.
However, the company found that in 2024, 43 percent of US work hours were not automatable, emphasizing the continued need for human empathy, creativity and judgment alongside these intelligent systems.
The study “completely defies the doomsday narrative,” Lee said.
“The future work isn’t about replacing man with machines,” Lee said. “Work in the future will be a partnership between people, agents and robots.”
Next up, Siemens Taiwan CEO Frank Grunert shared how industrial AI and digital twins are enabling companies to simulate, optimize and improve real-world systems before physical deployment.
NASA used a Siemens digital twin to run 8,000 simulations for a Mars rover landing, while Foxconn is using the simulation software to ramp up AI server production.
Siemens’ latest partnership with NVIDIA is now powering data centers, fusion reactors and the world’s first-ever fully AI-controlled factory in Erlangen, Germany, he said.
Meanwhile, Taiwan is poised for another “golden decade,” Grunert said.
“The real opportunity for Taiwan is ahead of us, which is not only building AI infrastructure but using AI in order to build things more efficient, faster,” he said, a prospect Taiwan is well prepared for as “one of the epicenters of manufacturing globally.”
From Taichung’s machine tool industry, to steel factories in Kaohsiung to Quanta’s autonomous driving developments in Taoyuan, Siemens’ technology is transforming Taiwan’s industries, traditionally “early adopters” of industrial AI technologies, he said.
“Consumer AI makes the headlines, industrial AI makes the impact,” Grunert said.
The final stage of the evolution is physical AI, Miller Chang, President of Advantech’s Embedded Sector, said.
He spoke on the importance of large-scale edge AI deployment in moving physical AI into the service industry, from factories to surgical robots and smart agriculture.
The main growth driving engine is edge AI’s low latency, enhanced data security and real-time performance, he added.
Kaliber’s partnership with Advantech has yielded surgical robots with millisecond-level-low-latency and high-resolution image processing, showcasing healthcare applications for physical AI at the edge.
Meanwhile, Techman is leveraging Advantech tools to build the robotic brains with the visual sensors and real-time inference capabilities needed to develop the “next-gen humanoid robot.”
Chang concludes the final presentation alongside a humanoid robot, which proudly presents Advantech’s latest WISE-Edge Developer Architecture (WEDA), a set of development packages to implement physical AI into industry.
The speakers presented a picture of the next chapter of AI: creating factories, infrastructure and collaborative workplaces that will power the global economy.
