Sunrise Robotics, a startup building modular industrial robotics and AI models that makes them simple to deploy in different environments, has emerged from stealth with $8.5 million in seed funding.

The investment round is being led by Plural, a London-based early stage venture capital firm formed by a group of prominent startup founders including Wise cofounder Taavet Hinrikus and SongKick cofounder Ian Hogarth. Venture capital firms Tapestry, Seedcamp, Tiny.vc and Prototype Capital also participated in the funding.

Sunrise, which is headquartered in Ljubljana, Slovenia, declined to comment on its valuation following the funding round.

The startup is trying to address an acute and worsening labor shortage in many European manufacturing firms, Tomaz Stolfa, its cofounder and CEO, said. These businesses currently represent 15% of Europe’s GDP and employ 32 million people. But close to a third of this existing European manufacturing workforce is set to retire in the coming decade and industrial companies are already saying they cannot find enough young workers to replace those who are leaving. Sunrise sees industrial robots being able to take over some of the manual cutting, welding, fastening, and bolting human workers currently perform on the production lines of these businesses.

Photo courtesy of Sunrise Robotics

The company can have its two-armed robots up-and-running on a new industrial production line in less than 10 weeks, Stolfa said, while it can take as long as eight months to deploy traditional industrial robots that have to be programmed on site.

The startup accomplishes this by using cameras to gather detailed three-dimensional data on the workstation where the robot will be deployed and also recording the steps a human worker currently takes to accomplish a task at that workstation. Sunrise uses this camera data to build what is essentially a digital twin of that workstation and trains AI models in a simulator that can control its robots to complete the task. Then it transfers this control software to its real robots.

Sunrise Robotics is not the only robot startup trying to use modern AI techniques and modular designs to make the delivery of robots for factories and warehouses much faster and more affordable. Paris-based Inbolt is also targeting industrial robotic arms, while Physical Intelligence is building “foundation models” that will enable any robotic arm designed to pick up and maniuplate a wide variety of objects.

Stolfa says that Sunrise’s software uses a combination of small AI models and conventional computer coding to control its robots. He said that as its robots master new skills, the time it should take to deploy them to future environments that demand similar skills should shorten considerably.

He also says that Sunrise’s decision to build standardized “cells”—as it calls its robotic workstations—makes it easier to train the robots for new tasks. The robot workstations are Sunrise’s own design, but are composed of mostly off-the-shelf parts, which makes them cheaper to build and maintain. “What we’ve done is we’ve productized the hardware,” he said.

Stolfa said one reason traditional industrial robots were expensive and time-consuming to deploy is that they were often designed specifically for one particular assembly line. This meant that only the largest manufacturing companies could afford to use them.

Sunrise, Stolfa said, is not targeting these businesses, such as major automakers. Instead, he said the company is going after the 60% of European manufacturers that are “high mix, low volume,” meaning that they produce a lot of different parts, but a relatively small number of finished products. He said Sunrise’s sweet spot was probably companies producing less than 100,000 parts each year, but that it could also work for those producing up to about 400,000 parts.

So far the company said it has signed letters of intent with about 10 customers, including those in supercar development, high-performance batteries, and consumer electronics manufacturing. Andrew Buss, the managing director at Asteelflash, an electronics manufacturer based in Bedford, England, that is an early Sunrise customer, said in a statement that the startup has helped it “adopt cutting-edge innovation at remarkable speed. Just a few months after initial data collection, we had a fully-trained, operational-intelligent robot up and running within hours of delivery.”  

Two of Sunrise’s three cofounders are experienced entrepreneurs, and all three spent time working in tech in Silicon Valley. Stolfa co-founded a number of previous companies—including the voice-over-internet company vox.io and also the messaging app builder Layer. Cofounder Marko Thaler, the company’s chief technology officer, previously founded Airnamics, which built AI brains for robots and drones. Meanwhile, Joe Perrott, Sunrise’s third cofounder and its chief commercial officer, was head of global program management at PCH International, which helps businesses build supply chains, including finding contract manufacturing partners. Its clients have included Apple, Amazon, Google, and Square.

The company currently employs 25 people in Ljubljana and working in a dozen locations across Europe. Stolfa said it plans to use its new funding to expand its team and ramp up production of its robot workstations.

This story was originally featured on Fortune.com

© Courtesy of Sunrise Robotics

Today’s most advanced AI models are relatively useful for lots of things—writing software code, research, summarizing complex documents, writing business correspondence, editing, generating images and music, role-playing human interactions, the list goes on. But relatively is the key word here. As anyone who uses these models soon discovers, they remain frustratingly error-prone and erratic. So how could anyone think that these systems could be used to run critical infrastructure, such as electrical grids, air traffic control, communications networks, or transportation systems?

Yet that is exactly what a project funded by the U.K.’s Advanced Research and Invention Agency (ARIA) is hoping to do. ARIA was designed to be somewhat similar to the U.S. Defense Advanced Research Projects Agency (DARPA), with government funding for moonshot research that has potential governmental or strategic applications. The £59 million ($80 million) ARIA project, called The Safeguarded AI Program, aims to find a way to combine AI “world-models” with mathematical proofs that could guarantee that the system’s outputs were valid.

David Dalrymple, the machine learning researcher who is leading the ARIA effort, told me that the idea was to use advanced AI models to create a “production facility” that would churn out domain-specific control algorithms for critical infrastructure. These algorithms would be mathematically tested to ensure that they meet the required performance specifications. If the control algorithms pass this test, the controllers—but not the frontier AI models that developed them—would be deployed to help run critical infrastructure more efficiently.

Dalrymple (who is known by his social media handle Davidad) gives the example of the U.K.’s electricity grid. The grid’s operator currently acknowledges that if it could balance supply-and-demand on the grid more optimally, it could save £3 billion ($4 billion) that it spends each year essentially paying to have excess generation capacity up-and-running to avoid the possibility of a sudden blackout, he says. Better control algorithms could reduce those costs.

Besides the energy sector, ARIA is also looking at applications in supply chain logistics, biopharmaceutical manufacturing, self-driving vehicles, clinical trial design, and electric vehicle battery management.

AI to develop new control algorithms

Frontier AI models may be reaching the point now where they may be able to automate algorithmic research and development, Davidad says. “The idea is, let’s take that capability and turn it to narrow AI R&D,” he tells me. Narrow AI usually refers to AI systems that are designed to perform one particular, narrowly-defined task at superhuman levels, rather than an AI system that can perform many different kinds of tasks.

The challenge, even with these narrow AI systems, is then coming up with mathematical proofs to guarantee that their outputs will always meet the required technical specification. There’s an entire field known as “formal verification” that involves mathematically proving that software will always provide valid outputs under given conditions—but it’s notoriously difficult to apply to neural network-based AI systems. “Verifying even a narrow AI system is something that’s very labor intensive in terms of a cognitive effort required,” Davidad says. “And so it hasn’t been worthwhile historically to do that work of verifying except for really, really specialized applications like passenger aviation autopilots or nuclear power plant control.”

This kind of formally-verified software won’t fail because a bug causes an erroneous output. They can sometimes break down because they encounter conditions that fall outside their design specifications—for instance a load balancing algorithm for an electrical grid might not be able to handle an extreme solar storm that shorts out all of the grid’s transformers simultaneously. But even then, the software is usually designed to “fail safe” and revert back to manual control.

ARIA is hoping to show that frontier AI modes can be used to do the laborious formal verification of the narrow AI controller as well as develop the controller in the first place.

But will AI models cheat the verification tests?

But this raises another challenge. There’s a growing body of evidence that frontier AI models are very good at “reward hacking”—essentially finding ways to cheat to accomplish a goal—as well as at lying to their users about what they’ve actually done. The AI safety non-profit METR (short for Model Evaluation & Threat Research) recently published a blog on all the ways OpenAI’s o3 model tried to cheat on various tasks.

ARIA says it is hoping to find a way around this issue too. “The frontier model needs to submit a proof certificate, which is something that is written in a formal language that we’re defining in another part of the program,” Davidad says. This “new language for proofs will hopefully be easy for frontier models to generate and then also easy for a deterministic, human audited algorithm to check.” ARIA has already awarded grants for work on this formal verification process.

Models for how this might work are starting to come into view. Google DeepMind recently developed an AI model called AlphaEvolve that is trained to search for new algorithms for applications such as managing data centers, designing new computer chips, and even figuring out ways to optimize the training of frontier AI models. Google DeepMind has also developed a system called AlphaProof that is trained to develop mathematical proofs and write them in a coding language called Lean that won’t run if the answer to the proof is incorrect.

ARIA is currently accepting applications from teams that want to run the core “AI production facility,” with the winner the £18 million grant to be announced on October 1. The facility, the location of which is yet to be determined, is supposed to be running by January 2026. ARIA is asking those applying to propose a new legal entity and governance structure for this facility. Davidad says ARIA does not want an existing university or a private company to run it. But the new organization, which might be a nonprofit, would partner with private entities in areas like energy, pharmaceuticals, and healthcare on specific controller algorithms. He said that in addition to the initial ARIA grant, the production facility could fund itself by charging industry for its work developing domain-specific algorithms.

It’s not clear if this plan will work. For every transformational DARPA project, many more fail. But ARIA’s bold bet here looks like one worth watching.

With that, here’s more AI news…

Jeremy Kahn
[email protected]
@jeremyakahn

Want to know more about how to use AI to transform your business? Interested in what AI will mean for the fate of companies, and countries? Why not join me in Singapore on July 22 and 23 for Fortune Brainstorm AI Singapore. We will dive deep into the latest on AI agents, examine the data center build out in Asia, and talk to top leaders from government, board rooms, and academia in the region and beyond. You can apply to attend here.

This story was originally featured on Fortune.com

© Milan Jaros—Bloomberg via Getty Images