OpenMind, a Silicon Valley-based startup, is not building the next physical robot. Instead, it is developing the software foundation that will power many of them.
The company is positioning itself as the “Android” of robotics. Its goal is to provide a flexible, hardware-agnostic operating system called OM1.
This system allows different robots, regardless of their form, to function and learn in ways that feel intuitive and human-like.
By focusing on the underlying software, OpenMind aims to unlock the true potential of humanoid and service robots.
These machines will not only perform tasks. They will also understand context, communicate, and improve over time.
Human-Level Intelligence
Robots have performed repetitive tasks for decades: industrial arms, automated cleaners, and warehouse bots are now common.
However, humanoid robots are entering new domains. They are being designed to live and work in homes, schools, and public spaces.
These roles demand more than basic programming; they require dynamic interaction with humans.
Machines must now learn, adapt, and respond with a level of intelligence that mimics human reasoning.
According to OpenMind founder Jan Liphardt, this transition requires a new approach.
“Machines are starting to interact with people in ways I’ve never seen before,” he explained. “This is no longer about automation. It’s about collaboration.”
Introducing OM1
OM1 is designed to solve this problem. It is open-source, customizable, and compatible with a wide range of robot types.
The system allows robots to interpret their environment, communicate with other machines, learn tasks without human instruction, and adapt quickly to changing scenarios
By building a software layer, OpenMind gives robots a shared language. Developers no longer need to build unique code for each machine.
Instead, they can build on a unified platform that grows smarter with use. This structure mirrors the success of Android.
Just as Android powers smartphones from different manufacturers, OM1 aims to support diverse robotic platforms.
FABRIC
Alongside OM1, OpenMind has also developed a protocol called FABRIC. This tool allows robots to verify identity, share information, and learn from each other instantly.
Liphardt described a compelling use case: one robot learns how to speak German, and using FABRIC, it shares that knowledge with another robot on the opposite side of the world.
This will occur without the need for human intervention. Such collaboration dramatically speeds up training and deployment.
Robots gain new skills through shared experiences. This type of learning mimics human networks- fast, scalable, and deeply connected.
“Humans have systems that let us trust, communicate, and coordinate,” Liphardt said. “Machines must build similar structures.”
Deploying Robots
OpenMind plans to ship its first fleet of OM1-powered robots by September 2025. These robots are quadrupeds, commonly referred to as robot dogs.
They will serve as real-world test units. “We expect users to come back with many requests,” Liphardt noted. “We want to iterate as fast as possible and address what matters most.”
This agile approach reflects OpenMind’s strategy. Rather than wait for perfection, the company releases early, listens carefully, and evolves rapidly.
Investor Confidence
OpenMind recently closed a $20 million funding round. The investment was led by Pantera Capital, with support from Ribbit, Coinbase Ventures, Pebblebed, and several angel investors.
This level of backing reflects growing investor confidence in robotic software. Hardware is still important, but scalable intelligence, the software layer, is where many see long-term value.
With this funding, OpenMind will expand its development team, run more field tests, and prepare OM1 for wider adoption.
