From Robotics to Physical AI: What Pittsburgh Robotics & AI Discovery Day Revealed About the Future

By William Powers, IQ Inc. Engineering Manager

Anyone who walked through Pittsburgh Robotics & AI Discovery Day this year expecting a traditional robotics show may have noticed a shift. There were still plenty of robots, from autonomous vehicles and drones to industrial automation and advanced manufacturing systems. But the conversation increasingly centered on the artificial intelligence inside the machines rather than the machines themselves.

Held September 16 at the David L. Lawrence Convention Center, the event brought together more than 250 exhibitors and thousands of attendees, including companies, researchers, investors and engineers working across robotics, AI, autonomy and advanced manufacturing. That mix points to something important: the future of robotics is being defined by the intelligence in the machine, not simply the machine itself.

For years, the robotics conversation focused on physical capability: whether a robot could move through an environment, manipulate an object, navigate safely or perform a repetitive manufacturing task. Those questions still matter, but AI is changing what comes next. Today's systems need to understand what is happening around them, interpret complex information, make decisions, adapt to changing conditions and interact with people.

That requires software and AI capabilities that go well beyond traditional robotics programming, and it is driving a convergence of computer vision, machine learning, edge computing, autonomy and increasingly sophisticated AI models. The term "physical AI" captures this evolution well. It describes AI systems that do more than generate information on a screen: they perceive and act in the physical world.

One of the most striking aspects of Discovery Day was the range of industries represented. Robotics and AI are no longer confined to the factory floor. The technologies on display have applications in transportation, defense and security, healthcare, logistics, infrastructure inspection, agriculture, energy, construction, and research and education. The event's programming reflected this breadth, with dedicated areas for robotics, AI and technology, advanced manufacturing, defense and career pathways.

That diversity matters because the hardest problems in robotics aren't necessarily about building a better robot. They're about deploying intelligent systems into complicated real-world environments. A factory floor is different from a hospital, a warehouse from a roadway, and a defense application from a consumer product. Each setting brings its own requirements for reliability, safety, cybersecurity, data management, regulatory compliance and integration with existing systems, which is where software engineering becomes essential.

Another trend that stood out was the growing importance of where AI processing happens. Robotic systems often can't depend on sending data to the cloud and waiting for a response, especially when they are navigating an environment, controlling machinery or making safety-critical decisions. Latency, connectivity and reliability all matter, so more intelligence is moving to the edge, closer to the sensors, machines and people using the system.

This creates its own engineering challenge. Even the most powerful AI model must run within the limits of real-world hardware, including constrained compute, limited power, imperfect sensors and unpredictable environments. Making all of that work together is as much an engineering problem as an AI problem.

Perhaps the most significant change this year was the emphasis on moving technology beyond the research stage. The Pittsburgh Robotics Network described the 2026 event as having a stronger focus on B2B connections, customer discovery, commercialization, investment, partnerships and talent.

Pittsburgh has long been known for world-class robotics research, but the next phase of the industry is about more than proving that something can be built. The harder questions are whether it can be deployed, integrated with existing systems, operated reliably at scale and maintained over its lifetime, and whether it can meet the safety and cybersecurity requirements of its environment. These are the questions where engineering disciplines begin to converge.

This year's event was especially meaningful for those of us at IQ, now part of Critical Software, because it was the first time we had a booth at Discovery Day. The response was encouraging. Our booth stayed busy throughout the event, and we met engineers, technology companies and organizations working on fascinating problems. Many of those conversations went beyond technology itself and focused on real engineering challenges: how to develop, integrate, test and deploy sophisticated software and AI systems where reliability matters.

Those discussions reinforced something I've seen repeatedly. Organizations don't need another technology demonstration. They need help turning promising technology into working systems.

Pittsburgh has a remarkably strong local ecosystem, but the challenges being addressed here are global. Connecting that innovation community with engineering expertise from around the world can help move technologies from concept to deployment, and potentially from Pittsburgh to markets far beyond it.

I don't think the future of robotics lies in building robots that simply do more things. It lies in building systems that understand more, adapt more and operate more intelligently in the real world. The robot provides the physical capability, AI provides perception, reasoning and decision-making, sensors supply the information, edge computing delivers responsiveness, and software engineering ties it all together.

There is one more piece that shouldn't be overlooked: trust. As intelligent systems become more autonomous, the engineering behind them has to become more disciplined, not less. Testing, verification, cybersecurity, safety, data quality, observability and maintainability all grow in importance when AI moves from generating answers to influencing physical actions. That may be one of the biggest opportunities in the industry. The winners won't be the organizations with the most impressive demonstrations, but those that can turn demonstrations into reliable, secure and maintainable systems that solve real problems.

Walking around Discovery Day, it was hard not to be impressed by what is happening in Pittsburgh. The region's robotics ecosystem has grown far beyond a collection of robotics companies. According to the Pittsburgh Robotics Network, the regional cluster now includes more than 320 robotics, deep-tech and AI companies, spanning AI, autonomy, advanced manufacturing, defense, software, research, investment and entrepreneurship.

If this year's event was any indication, the next chapter will be less about robotics versus AI and more about what happens when the two become a single technology stack. It's an exciting place to be, and we're excited to be part of it.

We're continuing the conversation on robotics, AI and trustworthy engineering! Reach out to schedule a call with our team in Pittsburgh and Portugal at [email protected]