Antioch Gets $4.25 Million in Preseed Funding to Accelerate Testing of Autonomous Robots with Digital Twins

Robotics software testing startup Antioch Inc. says it’s turning science fiction into reality after raising $4.25 million in preseed funding today.

Antioch was founded earlier this year by a team of Stanford University alumni who had firsthand experience with the difficulties of testing autonomous software for robotics systems. Advances in AI promise to transform the capabilities of robots by making them more intelligent, allowing them to operate autonomously and safely alongside human workers in different environments.

But as promising as these systems are, they are extremely complex, and what works for one type of robot will not necessarily have the same results for another, because so many different types of robotic systems are being built today. For example, several companies are building humanoid robots, and there are dog-like robots for search and rescue missions, forklift-like robots for working in warehouses, robotic arms, and even drones that can be classified as robots. What that means is that the software that allows them to act autonomously must be customized for each type of robot, and that’s where the problems arise.

Testing new software in a robot usually requires setting up the physical machine, which takes a lot of time, energy and space. Every time the software is modified, engineers have to waste more time resetting the robot, or move it to a different environment to test different capabilities. It’s a long and arduous process that Antioch wants to help developers avoid.

The solution is a cloud simulation platform that allows teams to create a digital twin of any type of robot, allowing them to quickly validate any AI-powered system before deploying it in the real world. The goal is to eliminate the lengthy testing cycles that slow robot development. So instead of spending hours resetting a robot or moving it from one place to another, developers can just click a few buttons to prepare for the next test. Moreover, they can run thousands of tests in parallel by launching multiple digital twins at the same time.

Antioch co-founder Harry Mellsop previously worked at Tesla Inc., where he used internal simulation tools that allowed the company to develop and iterate robotics and other autonomous systems extremely quickly, and now he wants to bring the same capabilities to other companies. “Right now, the way companies train AI for robots is absurdly manual,” he said.

According to Mellsop, some of the robotics companies he spoke to have rented Airbnbs just so they could test domestic robots, while others have spent millions of dollars building mock warehouses to simulate the working environment of their robots. “(Testing) is slow, expensive and wildly inefficient,” he said. “With Antioch you can do all that virtually.”

Mellsop’s co-founders all have similar experience. For example, former Google DeepMind engineer Colton Swingle led large-scale validation projects that simulated different types of infrastructure to test software products. Collin Schlager, meanwhile, worked for Meta Platforms Inc.’s Reality Labs unit, where he was involved in building virtual simulations for product testing.

Antioch’s fourth co-founder is Alex Langshur, who was previously head of product at a startup called Deep Gray Research. He argues that the ability to simulate robotics is also critical to U.S. national security. He said the country’s manufacturing capabilities have been systematically eroded over the past four decades due to endless offshoring. Under President Donald Trump’s administration, the country is now trying to reindustrialize as a national security imperative, but the problem is that companies don’t want to pay the excessive labor costs associated with domestic production.

“The only economically viable way to reindustrialize is to expand our workforce with robotics and automation,” Langshur said. “Efficient, scalable testing to ensure robots are safe and effective is now the rate-limiting step to making this happen, which is why we decided Antioch needed to be built.”

The startup is operating in “stealth mode” so far, but has already secured a number of customers, including many of the country’s top robotics and autonomous software companies, and even some Fortune 500 companies, the company says. With today’s funding, it can make its software available to everyone before the end of the year.

The round was led by A* and saw participation from a host of other backers, including Abstract Ventures, MaC Venture Capital, Box Group and New Zealand-based Icehouse Ventures, as well as angel investors such as Foxglove Technologies Inc. Chief Executive Adrian Macneil and Palantir Technologies Inc. Chief Technology Officer Shyam Sankar.

A* General Partner Bennett Siegel said Antioch’s platform is key to unlocking the science fiction movie vision of robots becoming part of everyday life. “Companies currently spend hundreds of millions a year to have a horse in the AI ​​robotics race, but technology like Antioch’s can significantly reduce the hurdles, giving more companies the opportunity to innovate,” he said.

Photo: Antioch