Starting a company is hard enough without being taken to court by one of the most aggressive legal teams in the technology industry. Jay Li, founder of robotics startup Proception, knows that firsthand. His company was barely off the ground when Tesla filed a lawsuit against him, accusing him of stealing trade secrets from the automaker's Optimus humanoid robot program, where he had served as a technical lead. Despite the pressure, Li kept the company moving. Earlier this month, Tesla dismissed the lawsuit following a settlement, and Proception is now moving forward with new funding and its first product shipments.
"I think it's kind of like a resilience test, or pressure test," Li said in a recent interview. "People say that what doesn't kill you makes you stronger, right?"
Tesla did not respond to requests for comment on the matter.
An $11 Million Seed Round to Build the Robot Hand of the Future
With the legal dispute behind him, Li is now focused on what he describes as one of the most difficult unsolved problems in robotics: building a robotic hand that can truly replicate the dexterity of a human hand. On Monday, Proception announced it has closed an $11 million seed funding round led by First Round Capital, with participation from Y Combinator and early-stage investment firm BoxGroup.
Alongside the funding announcement, Proception confirmed it is shipping the first batch of its high-dexterity robotic hand to researchers and robotics companies, while simultaneously opening up orders to a broader customer base. Li's vision is for Proception to become the go-to supplier of robotic hands for companies that do not have the time or internal resources to develop what the industry refers to as "dexterous manipulation" on their own.
A Gap in the Robotics Boom
The robotics sector has attracted enormous capital and media attention over the past several years, yet Li argues that a critical component has been consistently underfunded and underappreciated: the hand. Despite all the progress in locomotion, perception, and general robot design, making robotic hands perform at a human level remains an elusive goal.
Interestingly, one of the most prominent voices raising this concern has been Li's former employer. Tesla CEO Elon Musk has publicly stated that robotic hands represent one of the most significant unsolved engineering challenges in the field. Musk has maintained that Tesla's Optimus robots could begin working in factory environments within a few years, but the broader scientific community is considerably more cautious about timelines.
Kevin Lynch, director of Northwestern University's Center for Robotics and Biosystems, stated in comments reported last year that his team estimates it will take roughly a decade before robotic hands are functional and useful enough to replicate meaningful human tasks. That assessment reflects the general consensus among robotics researchers.
Li believes Proception can move significantly faster than that, and the key to doing so, he says, lies in how the company collects and uses training data.
Rethinking How Robots Learn to Use Their Hands
The dominant method for training humanoid robots today involves teleoperation. In this approach, a human operator wears a virtual reality headset and physically manipulates objects from the robot's perspective, allowing the robot to observe and learn from those movements. While widely used, this method has notable limitations.
According to Li, one of the most significant drawbacks is that the human teleoperator receives no tactile feedback from the objects the robot is interacting with. The operator cannot feel what the robot is touching, which limits the richness of the data being captured. Additionally, this approach is constrained by how many physical robots a company has available at any given time - scaling it up requires a proportional increase in hardware.
Proception's approach is different. The company has developed a sensor-laden glove that human testers wear, paired with a headset, to capture detailed data about how a human hand interacts with objects - all without requiring a robot to be present in the process. According to the company, this method enables the collection of "human hand interaction data without requiring a robot in the loop."
The same glove technology is also integrated directly into the robotic hand that Proception is developing, functioning as a sensor-rich artificial skin. The hand features 22 degrees of freedom and multiple joints per finger, which together enable a wide range of dexterous movements. This design allows Proception and its customers to gather more precise, task-specific data that can help the robotic hand more closely replicate human-level manipulation.
Li believes this dual focus on advanced hardware and scalable data collection is what sets Proception apart from competitors who tend to prioritize one at the expense of the other.
"You need both hardware and data, and those need to come hand-in-hand to get dexterous manipulation to work. A lot of companies solely focus on hardware, or hardware plus non-scalable data collection. We're working on highly dexterous hardware plus highly scalable data. We believe that's a key combination to solve this problem."
Why First Round Capital Backed the Company
Bill Trenchard, a partner at First Round Capital who led the investment in Proception, echoed Li's assessment of the market opportunity. He described dexterous manipulation as a critical and often overlooked piece of the broader humanoid robotics story, and expressed confidence that Proception is positioned to lead in that area.
"We think they will have the best hand in the market, maybe the most sophisticated hand today, and the underlying data and models to support that," Trenchard said. "Dexterous manipulation is a very, very, very important part of the whole humanoid story going forward, and as many people have said, it's sort of the last mile of getting these robots to be truly performant."
Trenchard also made a point of noting how Li handled the legal challenge from Tesla while simultaneously trying to build a company. Rather than viewing the lawsuit as a red flag, Trenchard said he was impressed by how Li managed the situation with transparency and composure.
"He was very upfront with us when this came out, and I think the team did an amazing job of keeping their heads down," Trenchard said. "Jay's a very strong leader."
Looking Ahead - Including Back Toward Tesla
Now that the lawsuit has been resolved and funding is secured, Li is focused on scaling Proception's operations and expanding its customer base. He is confident that the company's approach to combining high-performance hardware with scalable data collection will prove out in the market over the coming years.
Li also noted, with a degree of confidence that might raise eyebrows, that having faced down what he described as Tesla's aggressive litigation team, he would not be surprised if his former employer eventually comes to Proception as a customer once the startup establishes itself as the leading supplier of robotic hands.
"I think it will happen," he said.
Whether or not that prediction proves accurate, Proception's emergence as a funded, product-shipping company - despite the legal turbulence of its early months - marks a notable development in the race to solve one of robotics' most persistent engineering challenges.
