Hyderabad powers Hexagon’s global push Into physical AI

New Delhi: Much of the software intelligence powering Swedish industrial tech firm Hexagon is being built in Hyderabad. The company’s largest R&D centre is in the city, employing around 2,200 people — significantly more than in Switzerland, one of its traditional engineering strongholds. The firm clocked over $6 billion in revenue in fiscal 2025.

The Hyderabad teams develop the software backbone connecting Hexagon’s hardware sensors to cloud platforms, processing vast volumes of geospatial and manufacturing data to deliver what CTO Burkhard Boeckem calls “a complete value chain”. Engineers work across divisions — from geospatial mapping to manufacturing intelligence — with strong cross-pollination. Core technologies are adapted across industries, accelerating innovation. This convergence has even pushed Hexagon into medical aesthetics, where scanning systems originally built for terrain and buildings now analyse the human face.




For over two decades, Hexagon has built a bridge between the physical and digital worlds. Known for precision measurement tools — from laser scanners to airborne sensors — the company sees itself at the forefront of what Boeckem calls the “age of intelligence”.

“Hexagon has been known for the last 20 years to take the physical world with all our measurement technology and transfer it into a digital twin,” he told TOI on the sidelines of the India AI Summit in New Delhi. A digital twin is a highly accurate virtual replica of a real object, building or even an entire city. “Once the world is machine readable, that is the best base for artificial intelligence.”

Hexagon’s tools scan cities from aircraft, map roads and buildings, and inspect factory components with microscopic precision. The company began embedding machine learning into products as early as 2012. When generative AI entered the mainstream, customers began asking: if the world is machine readable, can we query it?

That question has practical implications. A fully mapped city can be analysed by AI to classify surfaces and calculate how much roof space is available for solar panels. Several cities already use such insights to guide sustainability efforts.

Three years ago, Hexagon deepened its push into “physical AI” through a partnership with Nvidia. The ambition is not just to analyse the world but to act within it. “If a robot can sense its surroundings, perceive them and reason what it wants to do, we take it back into the real world. That’s the essence of physical AI,” Boeckem said.

In June last year, Hexagon unveiled a humanoid robot designed for industrial tasks — focused on inspection, handling and manipulation. At a launch event in Las Vegas, four robots autonomously scanned a car door with millimetre precision. Early adopters include automotive and aerospace manufacturers. Companies such as Tesla use Hexagon’s inspection technologies in their plants.

The company’s industrial footprint is vast. “About 90% of all smartphones in the world are touching Hexagon technology,” Boeckem said. In car production, quality checks that once happened in separate inspection rooms now occur in real time. The goal is “lights-out manufacturing” — factories operating autonomously and making decisions when something goes wrong.

For Boeckem, precision is non-negotiable. AI models must be grounded in reality to avoid “dimensional hallucinations”. Accurate spatial data ensures digital insights can be safely deployed in the physical world.

“AI becomes truly powerful when it moves beyond the screen and into the real world,” he said. “If a robot can sense, perceive and reason using accurate measurement data, it can act with precision. That bridge between digital intelligence and physical execution is what we call physical AI.”