The company, which announced its funding today, is backed by Shine Capital, a16z Speedrun, Box Group, Carnegie Mellon University Endowment, Innovation Endeavors and SV Angel. Safeworld addresses a central tension in modern robotics: the shift from predictable, rule-based algorithms to generative AI models that can produce unexpected outputs.
“The safety challenge that we’re talking about is a combination of, one, really advanced generative AI probabilistic evals — how do you underwrite the risk of a probabilistic system? The second part that’s really hard is the trust part, and you need both to deploy a robot,” Zhao said.
Safeworld’s approach involves building digital versions of real-world environments using simulation platforms like Genesis and MuJoCo. The company inserts a simulation of the robot — driven by its actual software — and then runs thousands of scenarios where human models interact with it. Kyle Wong described a common use case: a blind corner in a factory where a robot needs to know its safe speed and stopping distance to avoid colliding with a worker, even one carrying boxes.
“Tripping and falling is also a good example of something that we do a lot of testing with the simulation,” Wong said. “Otherwise, you would have to go and trip and fall for the robot, which is like a hard thing to be doing all the time.”
The founders argue that robot makers will want an independent third-party to validate safety, even if they have internal testing tools, because it allows sharing of safety cases across competitors. Vishal Dugar, CTO of Gritt Robotics, which is partnering with Safeworld, noted the difficulty of testing these systems. Gritt builds AI brains for robots that install solar panels and aims to move into more complex construction tasks.
“A lot of people are underestimating one how hard some of these edge cases are going to be to solve,” Zhao said. “It is not the robot in the vacuum, in the demo, that we are worried about. It is the robot that is deployed at scale, with people who potentially never operated a robot before.”
a16z Speedrun partner Jonathan Lai said the time to build industry safety standards is now, while robots are still being designed and deployed. “By the time you have robots in households colliding with kids and causing safety incidents, that’s way too late,” he said.