DeepMind selects startup for robotics accelerator; raises $11.7M

What happened?

Acumino, a London-based startup, has been selected by DeepMind (Alphabet's artificial intelligence division) to join its robotics accelerator. Following this endorsement, the company announced the closing of an $11.7 million seed funding round. The investment was led by Plural, a European deep tech venture capital fund, and included participation from other undisclosed investors. According to Sifted, this is one of the largest seed rounds in the AI robotics space in Europe during 2025.

Why is it important?

This event is significant for several reasons. First, DeepMind, known for its AI breakthroughs like AlphaGo, AlphaFold, and more recently Gemini, is putting its stamp of approval on Acumino, suggesting the startup has promising technology. DeepMind launched its robotics accelerator in 2022 to support startups integrating advanced AI into robots, and Acumino is one of the first selected. This reinforces the trend of big tech investing in robotics startups as a way to diversify their AI portfolios, similar to Google's acquisition of Boston Dynamics in 2013 (though later sold) or Amazon's Alexa Fund. Second, the size of the seed round ($11.7M) is considerably high for an early stage, indicating investor confidence in the AI robotics market. For context, the average seed round in robotics in Europe in 2024 was around $2-3 million, according to PitchBook data. This positions Acumino to scale quickly.

What consequences will it have?

The capital injection will allow Acumino to accelerate development of its robotic platform, which aims to create robots capable of performing general tasks in unstructured environments like warehouses or factories. This could be a breakthrough compared to traditional industrial robots that require highly controlled environments. The startup focuses on manipulator robots that learn through demonstration and reinforcement, similar to approaches by companies like Covariant (which has raised over $200 million) or Robust.AI (founded by Gary Marcus). However, the partnership with DeepMind could lead to synergies in reinforcement learning and robotic simulation, areas where DeepMind is a world leader. For example, DeepMind has developed technologies like MuJoCo (a physics simulator) and applied reinforcement learning to complex robotic tasks. If Acumino integrates these advances, it could reduce robot training time and improve generalization. In market terms, generalist robotics has enormous disruptive potential: the industrial and service robot market is estimated to reach $74 billion by 2027 (according to MarketsandMarkets), and AI-driven automation could capture a significant share. However, the startup faces the challenge of proving its technology works in real-world applications beyond labs. Competition is fierce, with giants like Tesla (with its Optimus robot), Figure AI (backed by OpenAI and Microsoft), and Agility Robotics (with Digit) advancing rapidly.

What should readers know?

Readers should note that generalist robotics remains a huge technical challenge. Unlike industrial robots performing repetitive tasks in controlled environments, robots operating in unstructured settings must handle variability, uncertainty, and safety. Acumino competes with other startups like Covariant or Robust.AI, as well as research labs at big companies like Google, Amazon, and NVIDIA. Funding is a first step, but the startup must demonstrate concrete results in real applications. Additionally, DeepMind's investment is not an acquisition but a strategic collaboration. DeepMind has not disclosed exact terms, but Acumino likely gains access to research resources, simulators, and possibly talent. For investors, the risk is high: failure rates in robotics startups are elevated, and many have disappeared after initial rounds (like Rethink Robotics, which closed in 2018). However, backing from DeepMind and Plural reduces relative risk. Readers should also know that Acumino's team consists of researchers with experience in robotics and machine learning from institutions like Imperial College London, adding technical credibility.

"Robotics is entering a new era driven by artificial intelligence, and Acumino represents one of the most interesting bets in this space" — analyst at TheVortiq.

Additional context

DeepMind launched its robotics accelerator in 2022 to support startups integrating advanced AI into robots. Acumino is one of the first selected, reinforcing the trend of big tech investing in robotics startups to diversify AI portfolios. This move is part of a broader wave: in recent years, we've seen OpenAI invest in Figure AI, Amazon back Agility Robotics, and Google acquire several robotics startups. The key difference here is that DeepMind not only invests but offers access to its research ecosystem, which could accelerate Acumino's development. Moreover, the timing is relevant: generative AI has matured, and now the focus is on applying these advances to the physical world. AI robotics is seen as the next big leap, similar to what the cloud was for software. However, there is skepticism: some experts point out that generalist robotics is still far from commercially viable, and expectations may be inflated. History shows that robotics promises have been cyclical, with hype periods followed by "AI winters" (like in the 1970s and 1980s). But with DeepMind's backing and capital injection, Acumino has a unique opportunity to advance in this field. The next 12-18 months will be critical to see if they achieve technical milestones that justify the investment.