The robotics investment boom is attracting a new breed of venture capitalists who are betting that artificial intelligence will transform physical work as profoundly as it has transformed software. Venture capital investment in global robotics and physical AI has grown from around $4 billion in 2019 to $26 billion in 2025, according to PitchBook data, with companies in the space raising more than $23 billion so far this year.

For decades, venture investors chased software companies that could scale without factories, supply chains, or machines. Hardware was considered a liability. But as artificial intelligence moves from chatbots into the physical world, a cohort of seasoned investors is doubling down on robotics and what Nvidia CEO Jensen Huang calls "physical AI," a term describing AI systems that can act in the real world.

What's Driving the Robotics Investment Surge?

Several forces are converging to make robotics investment more attractive than ever. Sensors, cameras, actuators, and other hardware components have become cheaper and more capable. Artificial intelligence allows robots to move beyond rigid, preprogrammed behaviors and adapt to unpredictable environments. Talent from Tesla, Waymo, Amazon, and other hardware-building companies has learned how to deploy technology reliably in the real world. And labor shortages combined with geopolitical pressure to rebuild supply chains have made automation more urgent.

The influx of capital has also brought in what industry veterans call "hardware tourists," investors with little experience in the category who are drawn in by hype and may be underestimating how difficult it is to build machines that work reliably outside a controlled demo.

Who Are the Key Players Shaping the Robotics Boom?

Business Insider identified 22 investors, from established names to rising stars, who are shaping the robotics boom. Their bets span humanoids, autonomous vehicles, warehouse automation, defense robotics, and the software models that could define the next generation of machines. Here are some of the most influential:

• Aidan Madigan-Curtis at Eclipse: A former Apple and Samsara executive who leads the firm's investment in physical-world companies. She backed Verkada, a security camera and software company, and helped incubate Bedrock Robotics, a startup founded by former Waymo engineers automating heavy construction equipment. She also invested in Simbe Robotics, whose Tally robot scans store shelves to help retailers track inventory and pricing.
• Anand Agarwal at Bessemer Venture Partners: Spotted the promise of warehouse robots long before Amazon did. In 2004, he backed Kiva Systems, which pioneered warehouse automation using fleets of mobile robots. When Amazon bought Kiva in 2012 for $775 million, Bessemer was the company's only institutional investor. His recent bets include Vention, Gather AI, and Mind Robotics.
• Aydin Zuberi at Red Glass Ventures: Began investing in deep tech nearly two decades ago and founded Red Glass Ventures last year. His robotics investments include a foundation model startup, a robotics data training startup, and Foundry Robotics, which is developing robots to automate factory assembly.
• Jia Zhan at Confluence Ventures: Invested in Dyna Robotics, which is building foundation models for robots, and Skild AI, which is developing a general-purpose "brain" meant to work across different kinds of robots and tasks. He believes robotics is still in the stage where basic building blocks are being created.
• Jeremy Levine and Talia Goldberg at Bessemer Venture Partners: Leading the firm's robotics and physical AI investment strategy. They have argued that robotics is nearing its "GPT-2.5 moment," where models are getting better but the gap between demos and real-world deployment remains wide.

Matt Ocko, cofounder and managing partner at DCVC, noted that the robotics boom represents a fundamental shift in venture capital priorities. "Now the cool kids have arrived," Ocko said, referring to the influx of top-tier investors into the space.

How Are Researchers Advancing Embodied AI?

Beyond venture capital, academic researchers are making breakthroughs in how machines perceive and interact with the physical world. Two University of Maryland researchers, Levi Burner and Eadom Dessalene, have been selected for prestigious Intelligence Community Postdoctoral Research Fellowships to tackle fundamental challenges in embodied AI.

Burner, who earned his Ph.D. in electrical and computer engineering in Fall 2025, is working on predicting atmospheric turbulence using advanced imaging systems and artificial intelligence. His team combines single-photon avalanche diode (SPAD) cameras, which can detect individual photons and capture imagery at microsecond speeds, with light-field imaging, which captures multiple views of a scene simultaneously along different atmospheric paths.

"Atmospheric turbulence is one of the most significant barriers to obtaining clear imagery and reliable optical communication over long distances. By combining new sensing technologies with machine learning, we're exploring whether these distortions can be predicted before they occur, opening the door to entirely new capabilities in imaging and communication," said Christopher Metzler, an assistant professor of computer science at the University of Maryland Institute for Advanced Computer Studies.

Christopher Metzler, Assistant Professor of Computer Science, University of Maryland Institute for Advanced Computer Studies

Dessalene, who will complete his Ph.D. in computer science this summer, is focusing on helping artificial intelligence understand how humans physically interact with objects. His research combines egocentric, or first-person, video with electromyography (EMG) data collected from wrist-worn sensors. EMG measures the electrical activity produced by muscles, providing critical information about hand movements, contact, and force that may be invisible in video alone.

"Eadom's research tackles one of the most fundamental abilities of the human cognitive system, how we learn by watching others. What sets his work apart is the way it combines visual information with signals from the body itself. By capturing muscle activation alongside video, he is creating a technology with applications spanning healthcare, education, defense and robotics," explained Yiannis Aloimonos, a professor of computer science at the University of Maryland.

Yiannis Aloimonos, Professor of Computer Science, University of Maryland Institute for Advanced Computer Studies

What Real-World Tests Are Pushing Physical AI Forward?

Beyond the lab, physical AI is being tested in high-stakes competitions that reveal how well robots can perform in unpredictable environments. The Hitch Open Ping-Pong Embodied (HOPE) AI Challenge has been selected as an official competition of the Second World Humanoid Robot Games, which will run August 22 through 26, 2026, at the Ice Ribbon in Beijing.

Table tennis is brutally hard for robots. The ball moves fast, spins hard, and lands unpredictably, leaving only milliseconds to respond. The HOPE AI Challenge rules require full autonomy with no human in the loop. Each robot must track the ball, predict its trajectory and spin, choose a shot, plan its motion, coordinate its entire body, and correct errors in real time.

The inaugural World Humanoid Robot Games in 2025 drew 280 teams and more than 500 humanoid robots from 16 countries and regions, generating 1.33 billion views across media platforms and earning coverage from mainstream outlets in more than 80 countries. The Second World Humanoid Robot Games has expanded to five days around a simple theme: more autonomous, more dexterous, more practical.

"Unlike scripted routines or remote-controlled demos, the HOPE AI Challenge tests what a robot can do entirely on its own, the truest measure of physical AI in dexterous manipulation and embodied interaction," according to organizers of the Hitch Open platform.

Hitch Open, Intelligent Racing Inc.

What Should Investors and Entrepreneurs Know About the Robotics Boom?

While the robotics investment boom is real, experienced investors are sounding notes of caution. Aydin Zuberi, who has invested in deep tech for nearly two decades, warned that the space may be overheated. "My yellow flags are up. The space is overheated, overcrowded, and a bunch of noisy investments are being made. People will lose money. We are somewhere near the top of the hype curve," Zuberi said.

Agarwal, one of the most prominent robotics investors, has argued against the humanoid robot trend, writing in a Wall Street Journal op-ed that humanoids would prove to be a "parlor trick" with few practical uses. His portfolio reflects that thesis, focusing instead on warehouse drones and factory automation rather than human-shaped machines.

Despite the skepticism from some quarters, the fundamental drivers of the robotics boom remain strong. Cheaper hardware, more capable artificial intelligence, experienced talent, and urgent labor shortages are creating genuine opportunities for companies that can build machines that work reliably in the real world. The next few years will determine which investors and companies have correctly identified the most promising applications of physical AI, and which have chased hype into dead ends.