Safeworld raises over $12M to test robot safety with digital humans

In short: Safeworld has raised a seed round of more than $12 million to build simulation tests that check whether AI-powered robots could hurt people.

What happened

Safeworld is a new company founded by Dr. Ding Zhao of Carnegie Mellon University, startup executive Kyle Wong, and machine learning engineer Simo Rachidi. The company is coming out of “stealth,” meaning it had been working quietly before announcing itself publicly.

Safeworld says the big issue is that many newer robots are controlled by generative AI, which can be harder to predict than traditional software. In simple terms, it is like a robot learning to respond on the fly instead of following a fixed checklist, and that makes safety testing more complicated.

The company plans to test robots in simulated environments filled with realistic “digital humans.” Think of it like a flight simulator, but for robots that operate around people in factories, warehouses, and eventually homes. Safeworld says it can model situations like blind corners, people carrying boxes, or someone tripping and falling, then run thousands of test scenarios.

Safeworld’s seed round was led by Shine Capital and a16z Speedrun. Other investors include Box Group, the Carnegie Mellon University Endowment, Innovation Endeavors, and SV Angel.

Safeworld also says it is partnering with Gritt Robotics, a company building robots that work alongside humans on large solar farm construction sites.

Why it matters

As robots move into more everyday spaces, companies will need ways to show they are safe before accidents happen. Tools that simulate rare but realistic situations could help catch problems earlier, similar to how car makers test for crashes instead of waiting for real-world collisions.

Source: TechCrunch AI