On April 29th, 2025, Younite hosted an AI Transformation event focused on how artificial intelligence is reshaping industries.

The day was filled with engaging presentations and much discussion, going beyond the theoretical and deep into application of AI in practice. From robotics and simulation to operations and city planning, the conversations throughout the day were grounded in work already underway.

Henri Hyppönen, CIO of Younite opened the event with one of his many sharp observations:

This set the tone of the day. The opportunity is real, but so is the urgency. 

---

The KONE way of working

In an engaging panel discussion, Amy Chen, CIO of KONE, shared how her team is already navigating this shift. “The traditional production line won’t exist in 2030,” she said, noting that the company is not only working internal pilots, but with a wider AI community exploring practical use cases together.

Her key point: AI shouldn’t sit in R&D. For it to create value, it has to be integrated across operations, and that requires buy-in from leadership early on. “As fast as possible, move to execution,” she said.

Ilkka Lakaniemi from Aalto University picked up on the cultural dimension: the tech may be powerful, but its effectiveness hinges on trust and change readiness. That means leadership and deliberate work to make AI a part of how people operate, and create trust. 

An interesting discussion followed: How long did it take for you to start trusting driving directions given by Google Maps? When will you trust AI in directing how to do your daily work? 

---

Training AI for the Physical World

One of the most technically interesting parts of the event came from Tua Asplund of NVIDIA, who discussed the concept of Physical AI. 

“Physical AI will transform the world’s heavy industries - automating the world’s 10 million factories and two-hundred thousand warehouses. ​ Two of the largest opportunities are self-driving cars and generalist robots, or humanoids. These are the two form factors our constructed world was built for.“

Tua explained how virtual environments and tools like NVIDIA Omniverse are now being used to train these systems. Instead of learning through slow, expensive physical-world testing, AI agents can run millions of scenarios in high-fidelity digital spaces: learning how to move, adapt, and respond based on realistic physics.

She described a shift from Software 1.0 (where humans write code), to Software 2.0 (machine learning), to something that Jensen Huang highlighted recently at GTC: AI systems that perceive and act, trained in environments built for complexity. “Everything that moves, will be robotic someday, and that will be soon.”

This is especially relevant in industrial and logistics contexts, where downtime is expensive and precision matters. And while these systems are still maturing, they’re already being piloted in meaningful ways.

---

Digital Twins with Operational Value

So how to prepare for the age of Physica AI? Digital twins are already a proven tool. Throughout the day, we saw examples of how digital twin technologies are being used:

• Göteborg & Co, in partnership with Younite, is using digital twins to model visitor experiences and test design concepts before physical rollout.
  
  
• Scania’s work with simulation tools for aerodynamic design illustrated how even niche applications can generate measurable impact.
  
  
• Efeso and Younite showcased how simulation can drive efficiency in industrial operations towards “zero waste”.

These aren’t experiments, they are operational tools being used today. 

What used to feel speculative is now part of product roadmaps, industrial strategy, and day-to-day operations. AI is not on the sidelines, it’s moving into the center of how work gets done, also in the physical world. 

---

https://younite.ai/industrial-ai-event was organized as collaboration with Younite, NVIDIA, Combient Foundry and Business Finland in Helsinki 29th April.