Caterpillar (CAT) is expanding its use of artificial intelligence through a partnership with Nvidia (NVDA) aimed at bringing AI tools into construction equipment.

The company is piloting an AI assistive system in its mid-size Cat 306 CR Mini Excavator. The system, called Cat AI, is built on Nvidia’s Jetson Thor physical AI platform and is being demonstrated at the Consumer Electronics Show.

Caterpillar said the assistant is designed to support machine operators while they work. It can respond to questions, provide access to technical resources, share safety guidance, and help schedule service for equipment.

The system relies on multiple AI agents and is intended to operate directly within the machine. Caterpillar said this approach allows operators to receive information without stepping away from the job site.

The company said a key benefit of the technology is the data collected from machines in the field. Caterpillar’s equipment sends roughly 2,000 messages per second back to the company. That data is used to generate insights and improve performance.

Caterpillar is also testing digital twins of construction sites using Nvidia’s Omniverse simulation tools. These virtual models are being used to evaluate scheduling scenarios and estimate material requirements before work begins.

The company already operates fully autonomous vehicles in its mining business. Caterpillar said the current pilots represent a step toward expanding automation across a broader range of equipment.

Caterpillar said the projects were selected because they address immediate customer needs and could be brought to market relatively quickly. The company also said the work establishes a technical foundation for future automation efforts.

For Nvidia, the collaboration aligns with its focus on physical AI, which applies artificial intelligence to machines operating in real-world environments. The company has positioned physical AI as a major growth area for its computing platforms.

At CES, Nvidia outlined its physical AI ecosystem, including AI models, simulation software, and developer tools. Company executives said the strategy applies to a wide range of machines, from robots and autonomous vehicles to construction equipment.