Caterpillar CDO: AI Coaching Operators as 40% Face Retirement

Caterpillar is racing to deploy artificial intelligence across construction sites as the industry confronts a workforce crisis that will see nearly 40 percent of current equipment operators retire by 2031.

Ogi Redzic, who leads Cat Digital—a 3,000-person division building the software behind Caterpillar's heavy equipment—told Observer that autonomous tools and AI coaching systems represent a vital bridge across this labor gap. Construction jobs now take twice as long to fill as they did before COVID, and approximately 40,000 technicians retire in the U.S. annually.

"Humans won't be out of the loop anytime soon," Redzic said, describing a transition where workers shift from manual operation to supervisory roles overseeing autonomous workflows.

From scattered systems to AI-ready infrastructure

Redzic joined Caterpillar in 2018 when former CEO Jim Umpleby recruited him from Renault-Nissan to consolidate the manufacturer's fragmented digital efforts. At the time, each business unit maintained its own customer data, applications and services—an approach that couldn't scale across a global enterprise serving multiple industries.

His team built Helios, a unified platform that now stores 18 petabytes of data and processes roughly 700 billion pipeline transactions monthly. That early investment in data structure proved prescient when generative AI emerged. Caterpillar brought its AI Assistant to market in under a year because the underlying data was already organized and accessible.

Voice coaching and task autonomy on job sites

The AI Assistant provides real-time voice coaching to less experienced operators, eliminating the need for a senior operator to sit alongside them during training. Caterpillar has operated autonomous mining equipment for decades, logging hundreds of millions of kilometers without safety incidents. Construction sites present greater complexity, so the company is starting with "task autonomy"—individual machines executing specific assigned jobs.

Within five to ten years, Redzic expects AI to coordinate multiple machines across entire construction sites. He noted that up to 40 percent of time on typical sites is unproductive due to delays, missing design files, or machines waiting on each other. An orchestration layer could measure progress and adjust workflows in real time.

AI transforms software development itself

The technology is also reshaping how Cat Digital builds products. Redzic said his team now deploys platform services entirely developed by AI—from initial specs through design, coding, testing and deployment. The division accomplishes three times the work with only 20 to 30 percent more staff.

This shift has changed hiring criteria. Cat Digital now prioritizes "AI-native" talent who instinctively understand how to guide and govern AI-generated systems rather than write code line by line. The company recruits heavily from top computer science programs, and new hires contribute immediately.

Why it matters

Construction's labor shortage isn't a distant threat—it's already doubling hiring timelines and will accelerate as retirements peak. Caterpillar's approach shows how industrial AI can augment rather than replace workers, using automation to fill gaps while upskilling existing operators. The company's early investment in unified data infrastructure demonstrates that AI readiness depends on foundational work done years before the technology matures.

The details were first reported by Sissi Cao at Observer.