Safety is a critical aspect that is often overshadowed in the race to deploy humanoid robots in workplaces such as warehouses and factories. As corporations like Amazon, Mercedes, and BMW unveil plans to integrate humanoid robots into their operations, discussions around worker safety frequently take a back seat to the buzz around innovation and productivity gains.
One Bay Area robotics company, Figure AI, is determined to change that narrative. On Tuesday, the company announced the creation of the Center for the Advancement of Humanoid Safety, a dedicated in-house division aimed at ensuring that safety becomes a cornerstone of humanoid robotics development. The initiative is spearheaded by Rob Gruendel, a former Amazon Robotics safety engineer, who shared the announcement in a LinkedIn post.
The Growing Need for Humanoid Safety Standards
Traditionally, factory and warehouse robotics addressed safety concerns by isolating heavy machinery in cages, creating a physical barrier between robots and humans. Over time, advancements in technology, such as Veo Robotics’ software solutions and Amazon’s proprietary safety vest, have enabled robots to operate closer to human workers. However, even with these advancements, companies like Amazon still rely on protective cages for many of their robotic systems.
The humanoid robot form factor introduces an entirely new layer of safety challenges. Unlike stationary or task-specific robots, humanoid robots are designed to work side by side with humans, navigating shared spaces autonomously. This poses risks, as their large metal bodies and freedom of movement could lead to injuries if proper safety measures are not in place.
While workplace safety organizations like the Occupational Safety and Health Administration (OSHA) have established guidelines for automation, there are currently no robotics-specific OSHA standards, let alone regulations tailored to humanoid robots. This regulatory gap leaves significant room for error in ensuring safe human-robot interactions.
Figure AI’s Approach to Humanoid Safety
Figure AI’s new Center for the Advancement of Humanoid Safety aims to address this gap by proactively certifying its robots to meet industrial safety standards. Gruendel noted that the company has partnered with an OSHA-recognized independent testing laboratory to certify key components of their robots, including batteries, functional safety control systems, and electrical systems.
Gruendel emphasized transparency and collaboration in their approach, stating, “We recognize that the general population is often at the mercy of the technology experts to judge whether an AI-controlled robot can be safe. We want to speak directly with our customers.”
The company plans to test and communicate the safety performance of its robots across multiple parameters, such as:
- Stability while stationary and in motion
- Detection of humans and animals (including pets)
- Safe AI behaviors and navigation systems to avoid injuries
In addition, Figure AI will publish quarterly updates detailing their safety testing procedures, successes, and areas for improvement. These reports will provide transparency and foster trust as humanoid robots become an integral part of workplaces and, eventually, homes.
Industry Perspectives on Safety
The creation of the safety center aligns with a growing industry-wide acknowledgment of the importance of safety in humanoid robotics. Melonee Wise, Chief Technology Officer of Agility Robotics and founder of Fetch Robotics, has been a vocal advocate for improving safety standards in humanoid robots.
“With any humanoid robot operating in this space, safety is not clear,” Wise said during a humanoids panel at last year’s Automate conference. She criticized the lack of industry standards, highlighting that many humanoid robots lack basic safety mechanisms, such as emergency stop functions. “It’s very clear that many of the companies are not interested in it,” she added.
Looking Toward the Future
The work being done at Figure AI’s Center for the Advancement of Humanoid Safety is a significant step toward addressing these concerns. As Gruendel put it, “We will listen to suggestions from our customers on ways to test the safety of Figure 02 and beyond. We will summarize our successes and failures.”
These efforts could pave the way for establishing industry-wide safety standards, ensuring that humanoid robots can coexist with humans without compromising safety. With plans to bring robots into homes on the horizon, this work has never been more timely.
Humanoid safety standards are poised to become increasingly essential as more companies push the boundaries of AI and robotics, promising a future where robots enhance both workplaces and daily life.
