Related Experiment Video
Updated: Oct 17, 2025
08:34
Automated Charting of the Visual Space of Housefly Compound Eyes
Published on: March 31, 2022
2.1K
Memristor-based biomimetic compound eye for real-time collision detection.
Yan Wang1,2, Yue Gong1, Shenming Huang1
1Institute of Microscale optoelectronics and College of Optoelectronic Engineering, Shenzhen University, 518060, Shenzhen, P. R. China.
Nature Communications
|October 14, 2021
Summary
Researchers developed a novel artificial lobula giant movement detector (LGMD) using memristor technology. This compact device mimics insect escape responses for efficient collision avoidance in robots and biomimetic eyes.
Area of Science:
- Neuroscience
- Materials Science
- Robotics
Background:
- The lobula giant movement detector (LGMD) neuron in insects enables efficient collision anticipation and avoidance.
- Existing VLSI-based LGMD neuron simulations are bulky, energy-inefficient, and complex.
- Compact, simple devices mimicking LGMD escape responses are yet to be realized.
Purpose of the Study:
- To develop a compact and energy-efficient artificial LGMD visual neuron.
- To mimic the non-monotonic escape response of the biological LGMD neuron.
- To demonstrate applications in robot navigation and biomimetic compound eyes.
Main Methods:
- Implementation of an artificial LGMD visual neuron using a light-mediated threshold switching memristor.
- Utilizing the formation and breaking of silver (Ag) conductive filaments for non-monotonic response.
- Demonstration of robot navigation with obstacle avoidance and wide field-of-view (FoV) biomimetic compound eyes.
Main Results:
- A novel memristor-based artificial LGMD visual neuron was successfully implemented.
- The device exhibits a non-monotonic response to light flow fields, mimicking escape behavior.
- Demonstrated capabilities in robot navigation with obstacle avoidance and wide FoV detection.
Conclusions:
- Memristor technology offers a compact and efficient platform for artificial LGMD neurons.
- The developed device successfully mimics biological escape responses.
- Potential applications include advanced robotics and biomimetic sensory systems.
