[Edaily Marketin YunJi Kim Reporter] A British startup that can teach robots new skills simply by showing them a human performing a task just once has successfully secured a large-scale investment. Expectations that this technology could dramatically reduce the time and cost required for robots to learn new tasks attracted investors’ interest. As competition in AI-powered robotics intensifies, substantial funds are flowing into early-stage companies that have developed learning methods different from conventional approaches.

According to Imperial College London and other sources on the 30th, London-based robotics AI startup “Machenta” recently secured $40 million (approximately 53.5 billion won) in seed funding from European venture capital firm Crienダム and U.S.-based Union Square Ventures, among others. This investment marks the largest seed funding round ever secured by a British robotics startup.

Machenta is a startup that originated last year from the Robotics Learning Lab at Imperial College London. The technology the company is developing is an AI model that enables a robot to replicate a task exactly after being shown how to perform it just once by a human.

Particularly noteworthy is the method used to teach the robots. Traditionally, robots are trained by recording their movements one by one and accumulating data to train the AI; however, Machenta pre-trains its AI using data generated in a virtual environment on a computer. This allows robots to learn new tasks without having to repeatedly demonstrate countless movements in a real-world setting.

This learning method is considered significant because it can resolve the time and cost issues that have long been cited as obstacles to the adoption of robots in industrial settings. Collecting data and retraining robots every time they are assigned a new task in factories or logistics warehouses requires significant time and expense; however, it is explained that Machenta’s technology, once commercialized, could reduce this burden. This also means it would eliminate the hassle of having to retrain robots every time they are assigned multiple tasks or when the work environment changes.

Investors have also taken note of Machenta’s approach, which aims to transform the very way robots learn. In fact, investment firms reportedly recognized that Machenta’s research team had been overcoming technical limitations in the field of robot learning for a long time, even before the company was founded. USV, which led this investment round, also made the investment with the expectation that simplifying the process of teaching robots new tasks would bring about significant changes in terms of both speed and cost when deploying robots in actual industrial settings.

However, there are still challenges to overcome before the technology can be applied in actual industrial settings. Currently, Machenta’s AI model can perform some tasks after just a single demonstration, but it still has limitations when it comes to precise manipulation or complex movements. The company is also focusing on expanding the range of applicable tasks through research and development rather than commercializing the technology immediately.

Meanwhile, this investment comes amid a surge of large-scale funding in the U.S. for startups—such as Physical Intelligence and SkilledAI—that are developing robots’ general-purpose learning capabilities. The ability of robots to quickly and efficiently learn new tasks—rather than simply the number of tasks they can perform—has emerged as a key factor in determining technological competitiveness.

Machenta plans to allocate the funds secured through this round to advancing its AI models and expanding its research team. Currently employing 11 people, the company plans to establish a new office and research lab in King’s Cross, London, and expand its computing infrastructure.