Global physical artificial intelligence companies raised $47.4 billion across 521 deals during the first half of 2026, according to Crunchbase data, nearly quadrupling the amount compared with the second half of 2025, when companies in the sector raised $12 billion across 470 deals. The figure also exceeds the $26.4 billion raised in the first half of 2025 across 436 deals by approximately 80%.
Crunchbase uses the term physical artificial intelligence to refer to technologies that connect software with hardware, sensors and systems operating in the real world. This includes robotics, autonomous vehicles, aviation and space, drones, industrial automation and sensors.
A Surge Led by Massive Deals
The historical comparison illustrates the scale of the shift: venture capital funds invested only $41.9 billion in physical artificial intelligence companies over the combined three-year period from 2022 to 2024, less than the money that flowed into the sector during the first six months of 2026 alone.
Waymo’s $16 billion round, announced in February, was the biggest factor behind the surge. Alphabet, Dragoneer Investment Group, DST Global and Sequoia Capital co-led the round, which valued Waymo, based in Mountain View, California, at $126 billion. The round represented nearly one-third of the sector’s total funding during the period.
Other large rounds included the $5 billion raised by defense technology company Anduril Industries in May at a valuation of $61 billion, double its valuation of $30.5 billion less than a year earlier. In March, San Diego-based Shield AI closed a $2 billion Series G round co-led by Advent International and JP Morgan Chase, raising the company’s valuation to $12.7 billion.
That same month, Saronic, an Austin-based defense technology company developing autonomous naval vessels, raised $1.75 billion in a Series D round led by Kleiner Perkins. The round brought the company’s total funding to approximately $2.6 billion and its valuation to $9.25 billion, more than double the level of its 2025 Series C round.
Funding Accompanies Major Exits
Capital activity was not limited to investment rounds. The sector recorded several listings and acquisitions in 2026, with activity more heavily concentrated in space, defense and drones than in robotics.
SpaceX was the biggest exception, raising $75 billion in its initial public offering in June at a valuation of $1.77 trillion. Space intelligence company HawkEye 360, based in Herndon, Virginia, also raised $416 million in its listing, while autonomous drone company Aevex, based in Arlington, Virginia, raised $320 million.
In acquisition deals, Mobileye bought Israeli startup Mentee Robotics, which specializes in humanoid robots, for approximately $900 million. Mobileye explicitly linked the deal to its move toward physical artificial intelligence.
Why Does This Shift Matter to Investors?
Ryan Ziegler, general partner at Edison Partners, believes physical artificial intelligence extends beyond robotics, humanoid robots, defense and foundation models. In his view, the field combines software, hardware, sensors, the Internet of Things and services for applications spanning manufacturing, supply chains, utilities, agriculture, transportation, government, and spatial and physical intelligence.
Ziegler notes that artificial intelligence’s ability to process data generated by these systems rapidly and at scale is increasing, while hardware costs are declining and access to hardware is improving. He cited the presence of LiDAR scanners in mobile phones, considering this an expansion of the ability to map objects and spaces.
For Edison Partners, opportunities stand out particularly in high-value sectors that have traditionally relied on analog processes. The fund is looking for applications whose returns can be measured through predictive maintenance, risk management, asset safety, security and autonomous operations. It also believes some of these companies may resemble vertically specialized software companies, with multiyear deployments and substantial revenue, while the combination of software, sensors and hardware can produce proprietary datasets that increase in value over time.
From Experimentation to Commercial Operation
Joe Fath, partner and head of growth at Eclipse Capital, said physical sectors remain capital-intensive, but technologies supporting artificial intelligence are changing the efficiency with which companies are built and scaled. He attributes this to declining technical barriers, an influx of experienced professionals and rising market demand.
The economics of building these companies have also improved over the past two years, according to investor views cited in the article. Computing capabilities and foundation models have become more widely available, reusable models and physics-based simulation have improved, more training data has become available, and sensor and hardware costs have declined. At the same time, some companies are moving toward revenue models that combine hardware sales with recurring revenue or usage- and outcome-based pricing.
In practice, this means hardware may serve as a channel for distributing software and data, rather than merely being a standalone product. Fath says these factors are directing funding toward energy, robotics, autonomy, semiconductors, computing and data center infrastructure, with greater emphasis on companies that can reach production stages, acquire customers and scale efficiently.
Eclipse Capital defines physical artificial intelligence as “intelligence embedded in systems that perceive, reason about and act in the real world,” while it generally avoids investing in standalone large language model providers. Fath believes value may be distributed across the sector’s layers, but the strongest competitive advantages may belong to companies that integrate multiple layers of the infrastructure and application stack.
The conclusion presented by investors is not that technical capabilities alone are sufficient, but that customers care about operational efficiency, reliability and revenue. Therefore, the practical test for newly funded companies will be their ability to turn artificial intelligence into systems that can be relied upon commercially, and then use the resulting data and infrastructure to expand into additional products.