SafeWorld从隐形到建立 部署机器人安全模拟技术

全文内容 Full Text

SafeWorld said its platform can work in a range of industries with a range of robot embodiments, including robots arms, humanoids, mobile robots, and more. | Source: SafeWorld SafeWorld this week emerged from stealth with $12.2 million in seed funding. The startup aims to make safety testing on AI-powered robots easier with its simulation technology. “Our point of view is that if you’re going to have more capable robots, it just makes sense that you need to have more capable ways to test and to validate those robots,” Kyle Wong, co-founder and CEO of SafeWorld, told The Robot Report . “Otherwise, you’re not going to deploy those robots.” “We already have been using robots for maybe 100 years,” said Ding Zhao, co-founder of SafeWorld and director of the Safe AI Lab at Carnegie Mellon University. “We started the company today because, when AI came, it actually brought automation to the next level. The robots, for the first time, were allowed to go out of the cage and share the same space with humans.” Zhao asserted that this level of collaboration requires a whole new level of understanding for the robot, which is where SafeWorld comes in. Shine Capital and a16z Speedrun co-led the funding round. It also included participation from Box Group, Carnegie Mellon University Endowment, Innovation Endeavors, SV Angel, and other leading venture capitalists and angel investors. Ultimately, SafeWorld wants to make it easier for robotics companies to build safe systems. The Oakland, Calif.-based company currently has customers in industrial, manufacturing , logistics , construction , and even some home use cases, Wong said. “Robotics companies themselves need to, of course, be accountable for safety,” said Wong. “The question, though, is more, ‘Should they build it from scratch every single time?’ We don’t think that makes sense.” SafeWorld said it gives enterprises a scalable way to test how robots behave around people in rare, dangerous, and unexpected situations, without putting anyone at risk. “These are safety scenarios that have not existed yet in the world,” Wong said. “So then, how do you train against something that there’s not a wide dataset for? It’s a big challenge.” “We are providing robot safety simulation software that robotics companies or enterprises that are building robots can use to better provide evidence around safety, to test different types of safety scenarios before they happen in the real world, or test things that are effectively too rare, too dangerous, or too expensive to test in the real world,” he added. Teams can build scenarios right in the browser from past incidents, safety standards, and robot logs, with no simulation expertise required. SafeWorld then runs the robot through thousands of variations with realistic, reactive human motion and measures safety performance. Because every software update and new environment can introduce new risks, teams can rerun these tests continuously, not just once before launch. The result is a safety record that engineering, safety, and operations leaders can stand behind, claimed SafeWorld. Wong noted that while many companies work with simulation, using it for safety brings new requirements. “Where that breaks down and gets different is really in the nuance of the feature and functionality,” he said. “You have to figure out what even is the testing criteria for safety. You obviously have different standards, but those standards are not all complete. Then you have different types of risk assessments.” For some robotics developers, it can be hard to tell when it’s time to start investing in safety. If you start too early, your system may change significantly before seeing the real world, rendering much of your work useless. But, if you start too late, you run the risk of delaying deployments. “We work with about half a dozen companies, and two-thirds of them are publicly traded companies,” Wong said. “I would say that if the timeline for your deployment i