Max Robotics — Global Service Robot Directory + US Certification

About

Yinhe Han is a leading researcher in energy-efficient hardware acceleration for robotics and artificial intelligence, with a focus on specialized architectures for motion planning, collision detection, and vision processing. His most influential work, "DeepBurning" (2016), with 208 citations, pioneered machine learning accelerator design for neural networks, enabling innovative applications in embedded vision and cyber-physical systems. Han's "Dadu" family of accelerators represents a comprehensive approach to robotics hardware: Dadu-P (2018) tackles real-time motion planning in dynamic environments, Dadu-CD (2020) introduces processing-in-memory for collision detection, and Dadu-Eye (2021) achieves 5.3 TOPS/W for high-accuracy stereo vision at 30 fps/1080p. His more recent contributions include accelerating DNN-based 3D point cloud processing for mobile computing (2019) and the Dadu-RBD accelerator for rigid body dynamics (2023). With over 330 total citations across his top papers, Han's work consistently addresses the critical bottleneck of real-time performance and energy efficiency in robotics. His 2025 work on KARMA, an augmented memory system for embodied AI agents, signals an expanding focus on long-horizon task execution in household robotics.