Integrated a pinhole depth camera model with real-time range mapping (0.5m-12m) and live telemetry into the humanoid simulation! 🤖👁️ Great for depth perception and obstacle tracking.
Built entirely from a mobile phone!
#Robotics#ComputerVision#DepthSensor#Simulation#AI
Testing the P9 Local Autonomy stack in simulation! 🤖🗺️ Featuring multi-waypoint pathing, prior maps + live LiDAR perception, and dynamic obstacle/geofence handling.
Built and tested entirely from a mobile phone! 📱
#Robotics#LiDAR#AutonomousNavigation#AI
Added LiDAR and tested real-time autonomous navigation using an A* planner in MuJoCo simulation! Watch the humanoid scan, avoid obstacles, and hit the target.
Built entirely from a mobile phone!
#Robotics#LiDAR#AutonomousNavigation#AI
Testing object manipulation sequences (walk, brake, align, grasp, & lift) in simulation! 🤖📦
Using these trajectories as motion references for reinforcement learning to teach humanoids complex tasks. Built entirely from a mobile phone! 📱
Humanoid military support robots.
Built whole-body control policies (walks, squats) for a humanoid robot using just a mobile phone!
Goal: Humanoid military support robots.
Open to connecting with robotics engineers & hardware collaborators!
#Robotics#Humanoid#AI
Trained a walking locomotion policy on the Unitree H2 model—entirely from my Android phone over 12 days.
No laptop, no high-end PC, just resourcefulness.
My goal: building military-grade humanoid robotics.
#Robotics#AI#BuildInPublic