Safeworld Established to Test the Safety of Generative AI Robots
Safeworld has been established to enhance the safety of generative AI-based robots, raising over $12 million in funding led by Shine Capital and a16z Speedrun.
With the widespread adoption of generative AI in robotics, a startup named Safeworld has been founded to address safety concerns. The company evaluates robot control systems through simulations featuring realistic human models.
Generative AI and Robot Safety
One of the biggest trends in robotics is handing over control to generative AI models. However, this architecture is not as predictable as traditional algorithms and raises questions regarding human safety.
Establishment of Safeworld
Dr. Ding Zhao, who directs the Safe AI laboratory at Carnegie Mellon University, senior startup executive Kyle Wong, and machine learning engineer Simo Rachidi joined forces to solve this problem by founding Safeworld.
Investment Round and Supporters
Safeworld has emerged from stealth to announce that it has raised over $12 million in seed funding, led by Shine Capital and a16z Speedrun. The investment round also saw participation from Box Group, Carnegie Mellon University Endowment, Innovation Endeavors, and SV Angel.
Simulation-Based Evaluation Method
The company's core expertise is evaluating robot control systems through simulations populated with realistic human models. Digital spaces are built using models such as Genesis or MuJoCo, where thousands of scenarios are tested.
Sectoral Collaborations
Vishal Dugar, CTO of Gritt Robotics, which collaborates with Safeworld, pointed out the difficulty of mathematically proving robot safety due to the diversity of humans in workplaces, stating that empirical validation is essential.