Funding for artificial intelligence applied to real-world settings is accelerating fast. A monthly column on Crunchbase called Interesting Startup Deals You May Have Missed recently spotlighted two companies that illustrate this trend: Greyparrot, which uses computer vision to analyze recycling plant waste, and Emesent, which helps robots navigate underground where GPS does not reach. The column's subtitle, AI For Everything From Recycling To Breathing To Winning Construction Bids, captures the breadth of the technology's reach, though only the recycling and underground navigation deals were detailed.

What You Need to Know

Investors are pouring money into physical AI companies that operate outside the data center. Greyparrot and Emesent represent two distinct applications: waste management and autonomous navigation in GPS-denied environments. Their raises come amid a broader funding surge: physical AI startups raised nearly $47.3 billion in the first half of 2026, up almost 80% year over year, per Crunchbase data. Both companies are scaling to meet demand from industries like recycling, mining and defense.

Greyparrot Turns Trash Into Data

London-based Greyparrot raised a £20.3 million ($27 million) Series B led by technology investor Omar Mir. The company installs AI-powered camera systems above conveyor belts in recycling plants. Its computer vision software identifies materials, products and brands in real time. This data helps operators recover valuable materials, improve sorting efficiency and comply with stricter recycling regulations.

The systems are deployed in more than 20 countries and have analyzed over 1 trillion waste objects. Customers include Waste Management and Veolia. Greyparrot also feeds its data into a platform called Deepnest, which consumer brands such as Unilever, L'Oréal and Kenvue use to understand what happens to their packaging after disposal. The platform aids redesigns and helps companies meet Extended Producer Responsibility rules in Canada and the EU. The fresh funding will expand the company's footprint in North America and Europe and support a goal of preventing more than 1 million tons of waste by 2030.

  • Funding source: Crunchbase noted the investment as part of its ongoing coverage. Related Crunchbase queries show strong investor interest in physical AI.
  • Technology scope: Greyparrot's cameras and computer vision make every discarded bottle, carton and wrapper a data point.
  • Regulatory driver: Extended Producer Responsibility rules push consumer brands to track packaging lifecycle.

Emesent Navigates Where GPS Fails

Australian startup Emesent secured $17 million in new funding, including a $10 million equity round and a $7 million venture debt facility. Backers include Main Sequence Technologies, QIC Ventures, Orion Resource Partners, Hostplus and NGS Super. The company plans to scale manufacturing and expand its AI autonomy and cloud mapping platforms.

Emesent is best known for Hovermap, a LiDAR scanning payload that mounts to drones, vehicles or backpacks. It creates detailed 3D maps of mines, tunnels and other hazardous environments. The company's focus, however, is increasingly software. Its Cortex AI platform enables robots to navigate autonomously without GPS. Its Aura cloud platform processes and analyzes spatial data. The technology is deployed at more than 200 mine sites worldwide. Emesent is also expanding into defense, critical infrastructure and construction sectors.

  • Hardware and software: Hovermap provides the LiDAR data; Cortex AI enables autonomous navigation in GPS-denied areas.
  • Industry reach: Over 200 mine sites already use the system; defense and infrastructure are next.
  • Physical AI trend: Emesent's raise reflects growing demand for robots that operate in environments humans cannot safely enter.

Why This Matters

Both Greyparrot and Emesent show that AI investors are shifting focus from screens to the physical world. The nearly 80% year-over-year increase in physical AI funding signals that industries like waste management, mining and defense see clear returns from automation. For consumers, that means more efficient recycling systems and safer extraction of resources. For regulators, it means better data to enforce environmental rules. For the companies themselves, the challenge now is scaling production and proving that these technologies can work reliably at global scale.