Anyone benchmarking inference latency for onboard VLA models?

Whole-body control, RL policies, VLA models, sim-to-real, ROS2, and the software stack that makes a humanoid actually walk and act.
joseph31
Posts: 54
Joined: Mon Feb 02, 2026 4:33 pm

Re: Anyone benchmarking inference latency for onboard VLA models?

Post by joseph31 »

Slightly off-topic, but related: Whole-body control (WBC) formulates locomotion and manipulation as a single optimization problem across all joints simultaneously, respecting contact constraints and task priorities - it's more general than ZMP-only approaches but is computationally heavier and harder to tune.
he/him
gary.tanaka2
Posts: 86
Joined: Fri Nov 07, 2025 2:35 pm

Re: Anyone benchmarking inference latency for onboard VLA models?

Post by gary.tanaka2 »

@joseph31 From hands-on experience, A lot of what reads as 'full autonomy' in public demos is closer to a mix of scripted state machines, teleoperation for the hardest sub-tasks, and autonomous execution for the easier, well-rehearsed parts - transparency about this mix varies a lot between companies.
Post Reply