Pittsburgh’s physical AI bet depends on keeping companies local

Machines need somewhere to learn

Physical AI is built for machines that sense their surroundings and act on them. Unlike software, it cannot be tested only against material gathered online: robots need to operate in the places where they are expected to work.

That makes Pittsburgh’s industrial setting part of its technology strategy. The region has steel mills, power plants and rail infrastructure, alongside research at Carnegie Mellon and the University of Pittsburgh. Carnegie Mellon founded the world’s first robotics institute and has spent decades helping researchers become entrepreneurs.

The local companies span several kinds of work:

In 2025, equipment and robotics accounted for 51.8% of the region’s venture investment. Pittsburgh Robotics Network says the area is home to more than 260 high-tech companies employing over 11,300 people.

The case for being close to the work

Physical AI takes longer to pay off than software, and the places best positioned to build it are not necessarily the places with the most capital. Boston and the San Francisco Bay Area have more money and at least comparable research strength, but Pittsburgh is closer to industrial customers and sites where full-scale machines can be tested.

The region made that argument in September with a coordinated program of about fifty events over nine days. Fifteen partner organizations brought their programs together for the first time. Pittsburgh Robotics Network filled an exhibition center with more than 250 participants, including a defense pavilion. Army, Air Force and Marine Corps representatives visited the exhibition.

Policy was part of the pitch, too. Over 48 hours, events featured the United States’ top federal official for vehicle safety, the governor and both of Pennsylvania’s senators. Governor Josh Shapiro called for federal AI limits in a keynote. The following morning, John Fetterman and Dave McCormick spoke by video against such limits. The Democrat and Republican took a position opposed to that of the Democratic governor. McCormick said the government should act “modestly.”

For companies building trucks and robots for regulated industries, the disagreement is practical. A software product can be released and updated; a driverless truck or hospital robot needs approval, while agencies are still working out the requirements. Pittsburgh companies would rather face one federal rulebook than fifty state-level sets of rules. Yet the region’s own leaders have not agreed on whether those rules should exist.

I think the policy debate makes Pittsburgh’s pitch more concrete, but it also exposes a gap: proximity to industrial customers does not guarantee influence over the rules governing their machines. The region can make a stronger case if it can show the technology working in the field.

Keeping the value in Pittsburgh

The region’s challenge is to build commercial infrastructure around its technical base. Joanna Doven, who leads the AI Strike Team that organized the summit, has pointed to the risk that innovation happens in Pittsburgh while companies elsewhere capture the value.

The investment figures show why that concern is more than rhetoric. Local companies raised a record $2.06 billion in institutional venture capital in 2025, but more than three-quarters went to four companies. Pittsburgh venture funds had $23.2 million in available capital, more than 80% less than in 2022. A handful of fast-growing companies can secure large rounds; the next tier needs access to capital too.

That is only one part of the buildout. Pilot projects need to become repeat orders so young robotics companies can count on revenue. They need manufacturing space and places to test machines, not more offices. And the workforce needs somewhere affordable to live: from 2019 to 2024, Allegheny County lost 44,000 apartments renting for less than $1,000 a month.

I’d judge the bet less by how many companies Pittsburgh can attract than by whether those companies keep growing there. The useful signals are growth rounds led by Pittsburgh investors, repeat orders from paying customers and headquarters that stay in the region. The factories and research are already part of the story; the unresolved question is whether the people building this technology can build their lives there, too.