The frontier of physical AI is unfolding in a warehouse in San Leandro, California, where a game of Jenga is taking on new significance.
That warehouse is occupied by Encord, a company that builds data tooling for training AI models. Andrew Ceja is a pilot—the company’s term for its robotic trainers—and he’s carefully pulling wooden blocks from a tottering tower while wearing a headset with a camera that tracks his gaze. This is fairly standard for collecting robot training data, but this headset also includes sensors that measure his brain waves as he disassembles the block tower.
Encord is one of a small but growing number of startups betting that the next major constraint on humanoid and warehouse robotics will not be model architecture, but rather the scarcity of real-world physical training data. Instead of merely helping robotics companies manage the data they already have, Encord is building a business around manufacturing the data they lack.
The brain wave headset Ceja wears was built by Zander Labs, a German neuroscience startup that believes measuring brain activity—to infer mental states like error detection, intent, and surprise—can create a richer dataset for training models. Encord’s collaboration with Zander is currently a trial; the goal is to build an initial brain wave-tagged dataset, test it with customer robotics models, and evaluate whether it actually improves performance before scaling up.
Lucas Gehrke, a Zander neuroscientist supervising the work, explains that the amount of brain activity during a given task offers clues for model builders trying to determine when to deploy their most computationally intensive models.
This represents the “bleeding edge” of efforts to solve the robotics data bottleneck, according to Vineeth Velmurugan, Encord’s head of robot learning. A veteran of OpenAI’s robotics lab and warehouse automation firm Berkshire Grey, Velmurugan joined Encord to build the company’s internal data-creation team.
Encord was founded to help companies building machine-vision applications annotate data and evaluate models. As their customers—Velmurugan says they work with many leading robotics firms, but he is not authorized to name them—began applying end-to-end learning to robotic manipulation tasks, executives realized they would have to produce training data themselves rather than simply manage it. “The data simply does not exist,” Velmurugan said.
The bet that generative AI can do for robots what it has done for chatbots keeps hitting this same wall. LLMs were trained on the text of the entire internet, and more. Finding equivalent raw materials to teach neural networks about physical manipulation is challenging: self-driving car companies collect their own data, but that is difficult to scale. Training from video can work, but it lacks the fidelity of real-world data. Velmurugan estimates it will take a dataset roughly five times the size of YouTube’s video corpus to break through—a scale that helps explain why data generation has become a business in its own right, not just a research problem.
Feed Your Egocentric Data Needs
Companies building robot brains are now turning to two main sources: “egocentric” video collected by workers wearing cameras—often augmented with additional camera angles and other metrics—and data collected from robots operated remotely. Encord does both, gathering egocentric data from several factories around the globe and using its San Leandro facility to experiment with new modalities, such as brain waves, or to collect datasets for fine-tuning specific skills.
When TechCrunch visited, pilots were using these methods to push the boundaries of what physical AI can learn. As we move toward 2026, the race to overcome the data scarcity bottleneck is intensifying, with brain wave integration and novel data generation techniques emerging as key differentiators in the field.
via TechCrunch AI