PhysicalAI
Collection
2 items • Updated • 1
AI & ML interests
Physical AI
Recent Activity
View all activity
Papers
Can Predicted Dynamics Exist in the Physical World?
Silent Failures in Physical AI: A Literature Review of Runtime Action Authorization for Autonomous Systems
barakorÂ
updated a
collection about 19 hours ago
barakorÂ
submitted a
paper to Daily Papers about 19 hours ago
barakorÂ
submitted a
paper to Daily Papers about 20 hours ago
Post
87
Hi everyone,
I just published a new blog post from State16 about runtime integrity for Physical AI systems. The core question we explore is simple:
Can a predicted trajectory physically exist before it reaches the controller?
In many autonomous systems, a model can output a trajectory with a very high confidence score, but that trajectory may still violate basic physical constraints such as motion feasibility, velocity, acceleration, continuity, or interaction with the environment.
In our first paper and validation work, we tested this idea on LeRobot Push from Hugging Face. The goal is to detect physically inadmissible AI-generated behaviors before execution, especially in robotics and autonomous systems where a “confident” prediction is not enough.
Would be glad to get thoughts, feedback, and suggestions from the community.
Read here: https://huggingface.co/blog/barakor/can-predicted-dynamics-exist-in-the-physical-world
STATE16
I just published a new blog post from State16 about runtime integrity for Physical AI systems. The core question we explore is simple:
Can a predicted trajectory physically exist before it reaches the controller?
In many autonomous systems, a model can output a trajectory with a very high confidence score, but that trajectory may still violate basic physical constraints such as motion feasibility, velocity, acceleration, continuity, or interaction with the environment.
In our first paper and validation work, we tested this idea on LeRobot Push from Hugging Face. The goal is to detect physically inadmissible AI-generated behaviors before execution, especially in robotics and autonomous systems where a “confident” prediction is not enough.
Would be glad to get thoughts, feedback, and suggestions from the community.
Read here: https://huggingface.co/blog/barakor/can-predicted-dynamics-exist-in-the-physical-world
Post
106
I’ve just published a new article on State16 about a question I think will become increasingly important for robotics, embodied AI, VLAs, and world models.
Physical AI is already moving from research demos into real robots, vehicles, drones, and industrial systems. But as these models begin to produce actions in the physical world, we need to ask a harder question:
What is the runtime guardrail layer that makes these systems auditable, certifiable, and operationally safe?
In the post, I discuss the rise of World Foundation Models, Vision-Language-Action models, closed-loop autonomy stacks, and the missing layer between black-box AI policies and physical actuation.
Would be very interested to hear how this community thinks about runtime evaluation, action authorization, physical consistency checks, and safety layers for embodied AI systems.
Read here:
https://state16.ai/articles/physical-ai-is-already-here
Would love feedback, criticism, and pointers to relevant models, datasets, or evaluation frameworks.
Physical AI is already moving from research demos into real robots, vehicles, drones, and industrial systems. But as these models begin to produce actions in the physical world, we need to ask a harder question:
What is the runtime guardrail layer that makes these systems auditable, certifiable, and operationally safe?
In the post, I discuss the rise of World Foundation Models, Vision-Language-Action models, closed-loop autonomy stacks, and the missing layer between black-box AI policies and physical actuation.
Would be very interested to hear how this community thinks about runtime evaluation, action authorization, physical consistency checks, and safety layers for embodied AI systems.
Read here:
https://state16.ai/articles/physical-ai-is-already-here
Would love feedback, criticism, and pointers to relevant models, datasets, or evaluation frameworks.
barakorÂ
published a
Space 18 days ago