π0.5: A VLA with open-world generalization

It seems robotics has advanced more in the last 3 years than the previous 20.

When you're operating your robot around humans, you want to be very confident it won't injure anyone. It'd be pretty bad if a bug in your code meant instead of putting the cast iron frying pan in the dishwasher, it sent it flying across the room.

One way of doing that is to write code with no bugs or unpredictable behaviour, a nigh-impossible feat - especially once you've got ML models in the mix.

Another option is to put a guard cage around your robot so nobody can enter pan-throwing distance without deactivating the robot first. But obviously that's not practical in a home environment.

Another option is just to go slowly all the time. The pan won't fly very far if the robot only moves 6 inches per second.

The primary bottleneck is typically the motion planning system that must continuously solve complex optimization problems to ensure safe trajectories while avoiding collisions in dynamic environments.

Part of it is that training of these VLAs currently happens on human teleop data which limits speed (both for safety reasons and because of actual physical speed constraints in the teleoperation pipeline).

Let’s see how it changes once these pipelines follow the LLM recipes to use more than just human data…

Not a PI employee, but diffusion policies are like diffusion models for image generation, they generate actions from noise in multiple steps. With current compute you can't run 100+Hz control loops with that kind of architecture.

Some combination of distillation, new architectures, faster compute, can eventually attack these problems. Historically as long as something in tech has been shown to be possible, speed has almost always been a non-issue in the years afterwards.

For now even getting a robot to understand what to do in the physical world is a major leap from before.

Precisely my thoughts.

Ignore the haters. This is hilarious

Off the shelf robots -- we've got our models running on dozen+ different robot types (and have this specific generalization demo working on multiple platforms too.)

The development here is primarily in the model. If someone invents the 'brains' a robot needs to do useful domestic tasks then there will suddenly be a lot of incentive to build the right body for it.

One thing I notice is that they specify that the robot has never seen the homes before, but certain objects, like the laundry baskets, are identical.

Doing your demo is significantly easier if you've already programmed/trained the robot to recognize the specific objects it has to interact with, even if those items are in different locations.