Galbot is prepping a new humanoid for the August robot showcases
From a technical standpoint, the "Games" aspect of WHRG26 is what actually interests me. We've seen plenty of pre-recorded demos that look polished but fail in real-time. Seeing a new Galbot unit in a competitive or task-oriented environment is the only way to verify if their LLM agent integration is actually handling edge cases or if it's just following a rigid script. The gap between "lab-perfect" and "world-ready" is where most humanoid startups currently struggle, especially when it comes to dynamic balance and fine motor skills.
If Galbot is pushing for a debut at these events, they are likely betting on a breakthrough in how the robot perceives and interacts with unstructured environments. Most of the current hype is around end-to-end learning, and it will be telling to see if this new hardware is designed to support more complex AI workflows or if it's an iterative update to their existing chassis.
For anyone following the humanoid race, August is shaping up to be a critical window. We're moving past the era of "look, it can walk" and into the era of "look, it can actually be useful." I'm looking for specific data on latency and success rates in the tasks they perform at the games. If they can maintain fluid movement while processing complex commands in real-time, it proves that the hardware is finally catching up to the intelligence of the models driving them.
It's also worth noting that these conferences often serve as the primary battleground for visibility among investors and partners. A strong showing at WRC26 could shift the narrative around which companies are actually leading in humanoid deployment versus those just producing high-quality marketing renders. I'll be watching to see if the Galbot reveal includes any specific details on their sensory suite or if they keep the internals a secret for now.