From Elon Musk to SoftBank’s Masayoshi Son, tech visionaries have long imagined a future where humanoid robots work alongside humans in factories, schools, and homes. To realize this vision, manufacturers must ensure these robots pose no harm to people.
To address public concerns, multiple robotics companies are collaborating to establish industry standards for safe humanoid robot deployment in industrial settings. However, the standardization process not only involves tackling unique technical challenges but has also exposed disagreements among experts over safety stringency.
**Key Points:** - Industry groups are leading efforts to develop humanoid robot safety standards. - Powering down malfunctioning humanoids may cause dangerous tipping, creating new risks. - Standards may cover emergency stop buttons and anti-tip safety requirements.
The urgency for standards is heightened by humanoids’ physical dimensions—weighing up to 150 lbs (68 kg) and standing 6 feet (1.8 meters) tall. A falling robot could cause severe injuries, especially if handling sharp objects. Viral videos of失控 humanoids on social media have further fueled safety debates.
Standard-setting bodies face unprecedented dilemmas. Traditional industrial robot safeguards—like emergency power cuts—may backfire with humanoids. Sudden shutdowns can trigger uncontrolled collapses, endangering nearby personnel.
Carol Franklin, Robotics Standards Director at the Association for Advancing Automation (A3), explains: “Historically, the last line of defense was hitting that big red e-stop to cut power.” A3, representing over 1,400 companies, oversees U.S. industrial robot standards for arms, mobile robots, and now humanoids.
“With humanoids,” she notes, “hitting that button in a crisis might no longer be the right move.”
Standardization is critical for manufacturers targeting mass production. Tesla, Boston Dynamics, and startups like Apptronik, Figure AI, Agility Robotics, and China’s Unitree aim to deploy over 1 million humanoids in factories and warehouses within five years.
Safe shutdown protocols remain a key pain point. At A3’s quarterly expert meetings, discussions repeatedly circle this issue. Some warn that adopting stricter international standards could stifle innovation.
During a June committee meeting in Eden Prairie, Minnesota, debates arose over aligning with an upcoming ISO专项 standard for humanoids. Draft proposals reportedly require limiting severe injury risks from tipping to under 1 incident per million operating hours—a threshold potentially demanding redesigns.
Attendee Rob Grundel, then Figure AI’s safety lead, suggested creating a competing, more lenient standard. Grundel was later terminated; his proposal’s alignment with Figure’s stance remains unclear. (He has since filed a whistleblower lawsuit alleging safety negligence, which Figure denies.)
Figure AI, now valued at $39 billion after three years, hasn’t publicly commented on standards but joins peers in calling for updates to address shutdown risks. A spokesperson stated: “Safety is paramount, but current certification lags behind technological progress.”
**Emergency Stop Mechanisms** Post-Grundel, the committee proposed interim U.S. guidelines, which Franklin says may eventually yield to ISO’s 2028正式标准. Some manufacturers are innovating preemptively: Agility Robotics’ Digit humanoid crouches upon detecting hazards and collapses fetal-like when powered off.
Boston Dynamics’ Matt Powers emphasizes standardizing e-stop placement—on robots versus peripheral zones—and reliability requirements for safety systems.
Though non-binding, standards are commercially vital, proving risk mitigation to clients. Franklin adds that non-compliance may trigger OSHA scrutiny and insurer audits.
Eric Nieve, CEO of warehouse robotics firm Plussvan, recalls how collaborative robots struggled pre-2014 until certifications like TÜV Nord’s for Universal Robots unlocked market trust—a precedent humanoid makers hope to replicate.