Just completed the Montreal Robotics Summer School 2025 🇨🇦

It was an exciting and enriching experience where we explored advanced topics in robot perception, SLAM, sim2real transfer, and deep learning-based control. During the hands-on sessions, we tested our solutions on the LoCoBot and the Unitree Go1 quadruped robot, and also worked with NVIDIA Isaac Lab for simulation and deployment.

Highlights

• Implemented policy learning for locomotion/control; explored CrossFormer-based architectures for sequence modeling.
• Built and validated behaviors in Gazebo simulation; transferred policies to real hardware.
• On LoCoBot: pipeline for pick-and-place (perception → planning → execution).
• On Unitree Go1: training/tuning of policies and evaluation on real robot.
• Used NVIDIA Isaac Lab for sim & toolchains (dataset generation, evaluation, deployment).

🔗 Program info: Montreal Robotics Summer School 2025 @ Mila

[Summer School Gallery]