Iván Hernández Dalas: IEEE study group publishes framework for humanoid standards

image of a humanoid robot with the document title above it.

The IEEE Humanoid Study Group published the final version of its findings earlier this week. The mission was to establish a framework for developing standards for humanoid robots. ​ You can download a full copy of the report here.

This framework focuses on addressing the unique risks and capabilities of humanoids to enable their safe and effective deployment across industrial, service, and public applications. ​ The group aimed to coordinate efforts among Standards Development Organizations (SDOs) to create standards in three interconnected areas:

  1. Classification: Developing a clear taxonomy to define humanoid robots, their physical capabilities, behavioral complexity, application domains, and humanoid-specific traits. ​ This classification serves as the foundation for identifying applicable standards and addressing gaps. ​
  2. Stability: Creating quantifiable stability metrics, test methods, and safety standards tailored to actively balancing robots. ​ This includes addressing dynamic balance, fall-response behaviors, and predictive risk modeling.
  3. Human-Robot Interaction (HRI): Establishing guidelines for safe and trustworthy interactions between humanoid robots and humans, including collaborative task safety, interpretable behavior, and user training to manage risks and perceptions.

The working group consisted of over 60 individuals, led by Aaron Prather, director, robotics & autonomous systems program at ASTM International. The working group met for over a year, researching the market, interviewing vendors and end users, analyzing the needs, and debating the output.

pie chart breaking down the types of contributors to the working group.

The representative market segments of the individuals in the IEEE Humanoid Working Group. Credit: The Robot Report

“The hardest part was aligning such a diverse set of voices, from industry leaders to academic researchers to regulators, around a fast-moving technology,” said Prather. “Everyone brought different priorities to the table, and our task was to distill that into a set of shared recommendations without losing nuance. It was like building a bridge while the landscape underneath kept shifting.”

When asked what happens next with the standards development, Prather told The Robot Report, “This report is a starting line, not a finish line. We now need to turn the recommendations into concrete standards, testing protocols, and certification pathways. The real missing piece is ensuring these standards evolve alongside the technology itself, so they don’t just keep up with humanoids, but guide their safe integration into human spaces.”

The document is not a standards definition, and that was never the goal. Before the various standards bodies can complete new standards specifically for humanoid robots, the gaps in the existing standards need to be identified.

RoboBusiness, a leading event for robotics developers produced by The Robot Report that takes place October 15-16 in Santa Clara, Calif, will have a major focus on humanoids. The event will feature Agility Robotics and NVIDIA discussing lessons learned from early humanoid deployments. The event’s humanoid track will also cover the latest enabling technologies and design techniques that are vital to the evolving needs of humanoids.


SITE AD for the 2025 RoboBusiness registration open.

Likewise, the market opportunity and potential use cases need to be quantified. It was the two primary goals of the working group. The expectation now is that the various SDOs will use this document as a starting point to map out their individual projects, assemble the teams, and begin working through the details.

One of the more fascinating, thought-provoking, and heavily debated sections of the document was the “Classification of the Humanoid Robots.” This section generated a lot of debate about the definition of the various category levels. In the end, the team arrived at a well-thought-out methodology to characterize the systems.

Prather expects it will take another 18 to 36 months for the standards development work to be completed and the ratified standards to be published.

The net result is that, regardless of the wishes of individual entrepreneurs or venture capitalists to accelerate the introduction and adoption of humanoid robots, the opportunity to deploy a humanoid robot in volume will not happen until 2027 at the earliest. While there may be market demand, the humanoid companies will not be able to safely deploy humanoid robots collaboratively around humans until the new standards work is completed and ratified.

classic crossing the chasm diagram overlayed with humanoid path.

In the classic Crossing the Chasm model, the humanoid market is currently finding traction with early adopters. To cross the chasm to wider adoption, the robots must be safe to operate around people in the “beachhead” use cases that will drive adoption by the early majority. Credit: The Robot Report

One of the key remaining gaps to full autonomy for humanoid robots remains the lack of training data for the artificial intelligence that operates the robots. The 2-3 year window that remains in the safety standards development will likely reduce the pressure to rush humanoid robots to market, and instead enable the development teams to make pragmatic and measured progress in system development.

The prudent response for every humanoid robot developer and investor is to ensure that specific product roadmaps and milestones for these startups are funded through 2027.

 

The post IEEE study group publishes framework for humanoid standards appeared first on The Robot Report.



View Source

Popular posts from this blog

Iván Hernández Dalas: 4 Show Floor Takeaways from CES 2019: Robots and Drones, Oh My!

[Ivan Hernandez Dalas] Mechatronics in Ghost in the Shell

Iván Hernández Dalas: Tacta Systems raises $75M to give robots a ‘smart nervous system’