Almost every organization aiming to implement artificial intelligence encounters a similar issue: integrating the technology into routine practices is challenging. Industrial giant Caterpillar has spent many years tackling a version of that dilemma in the physical realm, and now it’s leveraging that experience to integrate AI.

Caterpillar’s venture into autonomy initiated with mining, where labor shortages and dangerous conditions create a significant need for automation. Currently, it offers automated haul trucks, drilling equipment, underground loaders, dozers, remote-controlled construction machinery, and more. Additionally, it provides software command centers, fleet management, and remote terrain intelligence as part of its autonomous technology suite.

“We are now in a thrilling era where we can apply all that knowledge from mining and extend it to more dynamic environments like jobsites, quarries, and construction locations,” stated the company’s CTO, Jaime Mineart, to TechCrunch during the Ai4 conference in Las Vegas earlier this month.

The industrial behemoth is now extending AI applications more broadly, including in tools utilized by technicians and its internal team. One instance is the Cat AI Assistant, allowing field technicians beside a machine to use voice commands for accessing repair procedures, troubleshooting issues, and pinpointing necessary parts prior to starting a repair. Mineart noted that the tool is currently used by customers, operators, and technicians alike.

The assistant relies on Caterpillar’s exclusive data, drawn from insights produced by its connected machinery. Mineart mentioned that Caterpillar has approximately 1.6 million connected assets worldwide and over 16 petabytes of structured information.

The organization is also utilizing AI to enhance software for site scanning and creating digital twins in manufacturing to scrutinize operations, she added. Similar to nearly all companies, Caterpillar is employing AI across its business processes, as well as in software development. “We implement AI agents to modernize outdated code, generate and test new software, and detect faults early on,” Mineart stated.

However, Mineart emphasizes that constructing the technology is merely part of the hurdle, as deploying an autonomous machine differs from altering a site to incorporate AI. Organizations must also reconsider how personnel collaborates with the technology and how existing procedures need adjustment.

“The challenging aspect of autonomy and physical AI is integrating that technology into the customer jobsite and into the workflows,” she explained.

Mineart mentioned that the company relies on seasoned operators to assist in training AI systems, harnessing institutional knowledge accumulated over years. As machines grow more autonomous, some operators may transition from managing a single machine to overseeing multiple machines from a remote command center.

This transition, however, presents a new obstacle for Caterpillar: training its 118,000 workforce. Mineart indicated that the organization plans to invest $100 million over the next five years to educate its employees in AI, autonomy, and robotics.

This investment is likely aimed at helping the company capitalize on the wider surge in AI infrastructure, which is already benefiting its revenue. Caterpillar’s quarterly earnings hit a record $20.5 billion in the second quarter, buoyed by robust demand for power-generation equipment utilized in data centers. Its power-generation segment experienced a 72% increase in sales, reaching $3.10 billion, with CEO Joe Creed stating that “no one is slowing down” regarding the demand for cloud computing and generative AI infrastructure.