Kion with GXO project in California
At the NVIDIA technology conference GTC in San Jose (California), Kion is currently presenting an autonomous forklift for warehouse use and an AI-based, safety-certified person detection system for automated truck loading in a combined “ GXO ” project

Kion is presenting two industrial AI applications for realistic scenarios – specifically for logistics. The focus is on autonomous material flow in real-world warehouse environments and safety-certified personnel detection for automated truck loading. This, the company says, marks the next stage in the strategic collaboration between Kion, NVIDIA, and the software experts at Accenture. The goal is to consistently transfer AI and digital twins into the realities of supply chains, seamlessly integrating AI-supported perception, simulation, and fleet management.
Transformation into reality
These flagship projects mark the transition from purely simulation-based development to real-world deployment in ongoing operations. For this, KION utilizes NVIDIA's Omniverse platform and its MEGA simulation solution, as well as a system architecture with an AI-powered digital twin. This creates digital representations of customer warehouses, enabling the virtual commissioning of systems and vehicles even before their physical installation. This allows for realistic testing of various scenarios – including safety-critical ones – under a wide range of conditions.

"Our customers are facing an increasing labor shortage and simultaneously need to improve their operational efficiency. We offer concrete solutions for this with intelligent autonomous mobile robots and AI-supported cameras deployed in the supply chain."
Tangible added value
To this end, an autonomous, AI-supported forklift been put into operation at a GXO Logistics logistics center in Épinoy, France. GXO Logistics, which describes itself as the world's largest pure-play contract logistics provider, is a leader in technology-enabled logistics and offers outsourced supply chains, warehousing, and returns logistics for leading brands in 27 countries. The site, which currently operates more than 200 manual forklifts, serves as a pilot environment for Kion's physical AI solutions.
“We are translating advanced AI into tangible added value for our processes – and this pilot project helps us explore the possibilities for the future of the supply chain.”
Image: Kion/NVIDIA
Using AI-powered cameras on the ceiling and on the vehicle, it recognizes pallets and transports them to defined target positions. Crucially, the autonomous vehicle works side-by-side with warehouse employees and manually operated forklifts – entirely without human intervention.
Together with NVIDIA, Kion is also working on the certification of a functional safety solution based on AI-supported person detection for warehouse environments. A "Halos Foundation Model", which detects and locates people and industrial trucks. This model is specifically trained for intralogistics applications using Kion-specific data. The solution is designed for the automated loading of trucks – a use case that was already demonstrated at CeMAT in Shanghai in October 2025. Kion is relying on permanently installed cameras connected to an NVIDIA Edge AI platform. A proof of concept in real-world warehouse operations is planned for later this year.
Borderline cases included
For safety-critical applications, AI models must also be prepared for rare but plausible situations. Kion generates synthetic training data in virtual environments to cover so-called long-tail scenarios – events with a very low probability of occurrence that nevertheless need to be reliably controlled. The virtual warehouse environment makes it possible to simulate such situations realistically without disrupting ongoing operations or having to conduct costly physical tests.
TERMS FROM THE LI GLOSSARY
ForkliftlogisticsMaterial flowSystem architecture (logistics IT)Supply ChainSupply chainLI-NEWSLETTER
Every week, the latest logistics news from Switzerland, the EU and the rest of the world.
Subscribe now




























