Kion drives AI partnership for logistics applications

The plan is to be able to define ideal set-ups for new warehouses and continuously enhance existing facilities with Mega, an NVIDIA Omniverse blueprint for large-scale industrial digital twins.

This includes a digital twin powered by physical AI – AI models that embody principles and qualities of the physical world – to improve the performance of intelligent warehouses that operate with automated forklifts, smart cameras and the latest automation and robotics solutions.

Kion Group CEO Rob Smith says: “With NVIDIA’s AI leadership and Accenture’s expertise in digital technologies, we are reinventing warehouse automation.”

Managing complex environments requires a delicate balance of precision and adaptability—a challenge compounded by fluctuating demand and shifting inventory needs, says Kion.

Functionality includes optimising the number of robots, workers, and automation equipment. The digital twin provides a testing ground for all aspects of warehouse operations, including facility layouts, the behaviour of robot fleets, and the optimal number of workers and intelligent vehicles.

The digital twin also trains the warehouse robots to handle changing conditions such as demand, inventory fluctuation and layout changes.

The digital twin is integrated with KION’s warehouse management software, and assigns tasks such as moving goods from buffer zones to storage locations to virtual robots. 

For the CES showcase, Accenture developed an interactive application that tracks how digital robot fleet and facility planning and testing enhances warehouse key performance indicators, such as throughput, task completion time, safety incidents and error rates.

As a next step, the partners are working to integrate the digital twin with a fine-tuned vision language model to capture real time insights from warehouses, reducing the risk of bottlenecks, accidents and other unforeseen events, pairing cameras, robots and NVIDIA NIM, a set of services for deployment of foundation models to edge devices in the warehouse.