The Embodied AI

1X Technologies, a startup from Norway, is quickly becoming a leader in the world of humanoid robots. These robots aren’t just built for industrial settings—they’re designed to be versatile assistants in a range of environments, including homes, healthcare facilities, and workplaces. 1X Technologies’ robots, could become the helpful companions that assist with everyday tasks, from organizing spaces to supporting daily routines. These machines are envisioned as general-purpose assistants capable of handling tasks that make life easier, whether in a home setting, helping the elderly, or assisting with repetitive duties in other environments

How It All Began

1X started back in 2015 (originally called Halodi Robotics) with a focus on making robots that move in a way that’s similar to humans. To do that, they developed a special motor technology, which acts like human muscles, allowing their robots to move smoothly and perform tasks that require precise movements. Their first model, EVE, rolled out in 2016, and by 2022, an updated version, EVE 2, became the first commercialized hardware from the company. The latest and most advanced robot from 1X Technologies, NEO, was unveiled in 2024 and quickly gained recognition as one of the most promising robots for assisting with everyday tasks.

NEO: The Next Step for Humanoid Robots

NEO is where things get truly exciting. Unlike many traditional robots, NEO is a tendon-driven machine designed to closely mimic the movements of the human body. It’s not just about flexibility—1X Technologies has invested significant effort into ensuring that NEO’s movements are safe and compliant, making it well-suited for environments where it interacts closely with people. The robot’s lightweight design and soft exterior are part of this safety-first approach. Rather than simply avoiding collisions, NEO’s advanced software helps it manage interactions safely by slowing down or adjusting its movements when it encounters obstacles, ensuring it can operate smoothly alongside humans without posing risks.

NEO also learns in a special way. With the help of AI technology and foundation models, the robot can understand spoken instructions and learn from demonstrations. So, instead of being programmed to do one specific thing, NEO can be taught how to complete new tasks, like setting the table or tidying up, just with the help of a human operator using VR.

What Makes 1X’s Approach So Successful?

1X’s robots are ahead of the curve for two big reasons: how they learn and how they move.

1. Learning from Demonstrations
Instead of programming robots line by line, 1X uses what’s called “learning from demonstrations.” This means robots like NEO leverage humans to collect data to learn how to perform a tasks and mimic human behaviors. The company uses VR technology to control the robots and teach them to do things like loading the dishwasher. This learning process, which is somewhat similar to how generative AI models like ChatGPT learn from large amounts of text data, helps the robots adapt to new tasks without needing every single instruction spelled out​. However, instead of pushing for large scale data collection, the company approaches the need of large data problem from a different perspective. Their focus is on operationalizing the data collection pipelines in new environments, ensuring that the variability of the data collected is well-balanced and targeted for environment-specific skills rater than very broad and general AI. This approach allows the robot to learn efficiently, acquiring the general skills it needs to perform in different spaces and situations, without requiring exhaustive amounts of training data​

2. Embodiment
The second reason 1X is succeeding is the design of their robots’ hardware, particularly the tendon-driven system that closely mimics human muscles. This design gives robots like NEO a similar range of motion and degrees of freedom to humans, allowing them to move naturally and perform tasks with precision. The tendon-driven mechanism enables the successful transfer of human demonstrations to robotic actions, overcoming the common issue of embodiment transfer, where differences in body structure and control systems between humans and robots make it difficult for robots to replicate human movements. By closely aligning NEO’s morphology with that of a human, 1X has made it much easier for the robot to learn from human demonstrations and handle objects with the same fluidity and dexterity.

The Future of Humanoids

Humanoid robots like NEO represent a significant leap in robotics, offering real potential for performing a wide range of daily tasks, fully powered by AI technologies. While robots have already proven their value in industrial settings, NEO is designed to support us in more personal and interactive ways, seamlessly integrating into various environments—whether it’s homes, workplaces, or public spaces. As 1X continues to advance its AI and robotics technology, robots like NEO could soon become a common presence in our daily lives.

1X’s success has been fueled by strong support from investors like OpenAI, which invested significantly in the company as part of its recent $100 million Series B round​. OpenAI’s involvement highlights the growing importance of integrating advanced AI with robotics. This partnership allows 1X to create robots that not only mimic human movement but also understand and respond to language in real-time, pushing the boundaries of what is possible with AI and robotics.

Beyond the technological innovations, the societal impact of 1X’s work is immense. As populations age, there will be a growing need for assistance with daily tasks—especially for the elderly. Humanoid robots like NEO could provide critical support in homes, helping with everything from household chores to elderly care. Additionally, the ability to automate labor-intensive tasks in industries could ease the burden on workers and help address labor shortages. As 1X continues to scale, their work could play a pivotal role in reshaping how we think about work and care in the future.