SAM-ACE
SAM-ACE is an integrated perception and control solution for autonomous legged robot to quickly traverse various terrains and accomplish designated tasks (e.g., environmental mapping) with customized payloads (e.g., 3D camera) in challenging environments. SAM-ACE enables legged robots to effectively handle discontinuous events and deliver desired performance across a wide array of uncertain environments. SAM-ACE is self-upgraded, trusted/reliable, and easily extended to different legged robots under different uncertain environments. IFT has successfully developed a standalone 3D camera subsystem for simultaneous localization and mapping that can provide accurate 3D environment maps.
Download the SAM-ACE
DOWNLOAD
METHOD:SAM-ACE is new-generation intelligent learning-based model predictive control, including situation-aware hierarchical bi-level hybrid dynamic modeling, hierarchical hybrid robust adaptive control, and physics-guided and low-computational-cost hybrid learning
KEYWORDS: Autonomous legged robots, state estimation, model predictive control, bi-level hybrid model, hierarchical robust adaptive control, hierarchical hybrid learning, sensor fusion, hierarchical actor-critic algorithm
TECHNICAL SPECS: Traversing five body lengths incline of +/- 20 degrees with root-like obstructions and slippery surfaces at 0.3 body lengths per second)