Papers

2

Total Citations

22

H-Index

2

About

Zhehong Wang is at the forefront of energy-efficient, domain-specific system-on-chip (SoC) design for autonomous micro-robotics. His research centers on enabling fully-on-chip intelligence for vision and navigation, a critical challenge for tiny, power-constrained robots. Wang’s major contribution is the development of novel hybrid processing elements (PEs) that seamlessly handle both convolutional neural networks (CNNs) and classic, non-CNN vision tasks, achieving unprecedented efficiency. His flagship work, the “RoboVisio” SoC, fabricated in 22nm technology, integrates 2MB of embedded MRAM for retentive, fully-on-chip weight storage, eliminating the need for off-chip memory access. This design, detailed in his 2022 paper (16 citations), delivers a remarkable 3.5 TOPS/W, demonstrating a flexible architecture for micro-robotic vision. His subsequent 2024 article (6 citations) further validates this approach for autonomous navigation. By pushing the boundaries of near-sensor processing, Wang is paving the way for intelligent, self-contained micro-robots capable of complex tasks without cloud dependency, marking a significant leap in edge-AI hardware.

Research Focus

Key Achievements

2
H-Index
2
Papers
22
Total Citations
11
Avg Citations/Paper
🏆 Most Cited Paper
A 22nm 3.5TOPS/W Flexible Micro-Robotic Vision SoC with 2MB eMRAM for Fully-on-Chip Intelligence
16 citations · 2022
📈 Most Prolific Year: 2022 (1 Papers)
🤝 Key Collaborators: 9
🏛 Institutions: University of Michigan–Ann Arbor, Meta (United States)

Top Papers

  1. 1
  2. 2

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 13 days ago