Where humans and robots learn to work together.
Social interaction with an autonomous walking (kid-sized) robot has never been this advanced! Meet PAWO (Personal Autonomous Walking Optimizer), a next-generation bipedal robot powered by cutting-edge AI and reinforcement learning. Designed for seamless human-robot interaction, PAWO learns to walk, balance, and engage with people using the latest in physical AI and social robotics.
🦿 AI-Powered Walking – PAWO refines its movements using deep reinforcement learning, adapting to new environments just like a human.
🤖 Engaging Social Behaviors – Autonomy is not about control but coexistence. It responds to gestures, voice commands, making every interaction dynamic and lifelike.
🎨 Cool & Playful Design – With a friendly, eye-catching appearance, PAWO is designed to be fun and approachable, making it especially appealing to kids and even those who may be unsure about robots.
💡 Future of Robotics – From assisting in education to research in human-robot collaboration.
PAWO is paving the way for a new era of autonomous companions. Watch as PAWO interacts in real-world scenarios, demonstrating how AI-driven robotics can revolutionize the way we connect with machines.
Full sync between simulation model and real robot
Demonstration of capabilities of NomOS control system with full stack HIL
Introducing Orion WALK3R – the Wise Autonomous Laboratory Keeper, 3rd Gen Robot! A revolutionary, fully 3D-printed autonomous robot designed for affordability, adaptability, and advanced AI-driven interaction. Unlike traditional humanoid robots, Centaur WALK3R features a modular design, allowing it to seamlessly switch between a bipedal humanoid and an agile quadruped, making it one of the most versatile robots in its class.
🦿 Physical AI for Adaptive Movement – Learns walking, balancing, and human-machine interaction using cutting-edge reinforcement learning.
🖨️ Fully 3D-Printed & Affordable – The majority of its parts are 3D-printed, making it accessible and easy to modify or repair.
🎨 Friendly Sci-Fi Droid Aesthetic – Inspired by classic movie droids, its design is charming, approachable, and built to delight rather than intimidate.
Orion WALK3R is the future of versatile robotics, bridging the gap between research, practical applications, and social engagement.
Foundations: What Humanoids Robots Can (and Can't) Do
Goal: Align expectations, avoid unrealistic assumptions
Outcome: To understand limits, risks, and realistic use-cases
Navigation & Interaction in Indoor Environments
Goal: Show why "just moving around" is hard
Outcome: to see hidden complexity of deployment environments
Simulation & Digital Prototyping
Goal: Show how problems are solved before real robots
Outcome: to understand how to test ideas cheaply and safely
Control, Learning & Deployment Stack
Goal: Show how robots actually learn and operate
To understand how behavior is created and maintained
Nice shirt which help us to build cool robots.
Build your own #PAWO, it looks cool on the shelf.
Cite This
S. Vechet, J. Krejsa and K. -S. Chen, "Vertical Stabilization of Bipedal Walking Drone PAVO with Proximal Policy Optimization," 2024 21st International Conference on Mechatronics - Mechatronika (ME), Brno, Czech Republic, 2024, pp. 1-6, doi: 10.1109/ME61309.2024.10789752.
keywords: {Legged locomotion;Training;Mechatronics;Machine learning algorithms;Navigation;Reinforcement learning;Robots;Optimization;Faces;Drones;Machine Learning;Proximal Policy Optimization;Walking Robot;Biped Robot},
BibTex
@INPROCEEDINGS{10789752,
author={Vechet, Stanislav and Krejsa, Jiri and Chen, Kuo-Shen},
booktitle={2024 21st International Conference on Mechatronics - Mechatronika (ME)},
title={Vertical Stabilization of Bipedal Walking Drone PAVO with Proximal Policy Optimization},
year={2024},
volume={},
number={},
pages={1-6},
keywords={Legged locomotion;Training;Mechatronics;Machine learning algorithms;Navigation;Reinforcement learning;Robots;Optimization;Faces;Drones;Machine Learning;Proximal Policy Optimization;Walking Robot;Biped Robot},
doi={10.1109/ME61309.2024.10789752}}