ASUS Humanoid “Gallop” Reportedly Learns to Package Motherboard Accessories

humanoid.guide reported on Nov. 14, 2026, that ASUS’ humanoid “Gallop” is learning to package motherboard accessories.

The report points to an effort to evaluate factory applications as tasks that must meet quality and time requirements, rather than as isolated movements. However, the success rate and time required were reported as targets, not results showing actual productivity or the amount of worker intervention.

The key takeaway from the report is that the criteria for evaluating humanoids are shifting beyond whether they can demonstrate a movement to whether they can consistently perform a manufacturing step according to quality standards and time requirements. In manufacturing, adoption decisions depend on more than one-off successes: they also depend on repeatability when handling different parts and packaging, recovery from failures, and whether the robot can operate without a worker constantly by its side. The figures in the report are development targets and cannot be treated as achieved performance. This distinction is important to avoid confusing successful examples on video with readiness for deployment on the factory floor. The task was described as involving multiple accessories of different shapes and flexible packaging. The evaluation therefore appears to cover not only grasping, but also object recognition, handling variations and completion of the entire packaging process. However, the reported information does not establish actual operating time, failure rates, the frequency of worker intervention or the method of quality inspection. Targets and actual results need to be tracked separately. Integration with existing equipment and safety measures are also essential to decisions about industrial robot deployment. For a humanoid, working in an environment with workbenches designed for people may indicate the potential to replace tasks without major changes to equipment, but the extent of modifications or supervision actually required needs separate verification. Plans for customer trials in 2027 and commercial availability in the second half of 2028 were also reported, but plans do not mean results have been achieved. The scope of the trials, evaluation criteria, stability during extended operation and safety are the next issues to examine. See the original report for details.