Booster_T1 is a humanoid robotics platform provided by Aaltoes, designed for research, prototyping, and education. It features high-torque joints, an NVIDIA Jetson AGX Orin (≈200 TOPS), RGB-D cameras, a microphone array, and a 9-axis IMU, making it suitable for navigation, perception, and interaction tasks.
Its sensor suite supports advanced applications like SLAM (Simultaneous Localization and Mapping), obstacle avoidance, and gesture recognition. The humanoid form factor makes it valuable for human–robot interaction studies and social navigation research, while the Jetson AGX Orin enables real-time AI inference on-device.
To begin development, set up ROS2 and the Jetson software stack. You can connect via the companion mobile app or SSH to verify sensors and joint control. Proper calibration and system checks are essential before deploying autonomous behaviors.
Development typically builds on ROS2’s modular ecosystem. Start with Navigation2 for path planning and obstacle avoidance, and expand with Isaac ROS for GPU-accelerated perception. Custom nodes in Python or C++ can be added for specialized behaviors such as gesture-controlled operation or mixed reality telepresence (e.g., combined with Varjo XR headsets).
Research opportunities include human–robot interaction studies, digital twin integration, reinforcement learning for locomotion, and robotics education. The platform is also suitable for experimenting with sensor fusion, control strategies, and AI-driven decision-making in humanoid systems.
Operational safety is critical: Booster_T1’s mass and high-torque joints require defined safety zones, an accessible emergency stop, and supervision during use. Battery runtime is limited, so plan experiments around charging cycles or tethered operation for long sessions.
Successful development requires ROS2 proficiency, familiarity with control systems, computer vision, and machine learning concepts. Understanding safe operation procedures for human-scale robots is essential for effective and secure deployment.
Essential Resources:
- ROS2 Documentation
- Jetson Docs
- Navigation2
- Isaac ROS
Custom AI workstation with Tenstorrent Blackhole™ P100a accelerator, AMD Ryzen 9 9950X3D CPU, 64GB DDR5 RAM, and 2TB NVMe SSD....
The Tenstorrent Blackhole™ p100a is a PCIe accelerator card with 16 RISC-V cores and 28 GB of GDDR6 memory. Designed...
Compact edge AI module delivering 67 TOPS with Ampere GPU (1024 CUDA, 32 Tensor cores), 6-core Arm A78AE CPU, 8GB...