Analog-digital hybrid computing with SnS₂ memtransistor for low-powered sensor fusion

Area of Science:

• Materials Science
• Computer Engineering
• Robotics

Background:

• Intelligent drone flight control typically relies on conventional digital computing, which is power-intensive.
• Robust drone operation across diverse environments necessitates energy-efficient computing solutions to minimize battery drain and computational load.

Purpose of the Study:

• To demonstrate a low-power analog-digital hybrid computing platform for drone sensor fusion.
• To leverage SnS2 memtransistors for energy-efficient drone flight control algorithms.

Main Methods:

• Development of an analog Kalman filter circuit utilizing SnS2 memtransistors.
• Integration of the memtransistor-based Kalman filter for sensor fusion (gyroscope and accelerometer data).
• Experimental verification of the hybrid computing platform's performance and power efficiency.

Main Results:

• The analog Kalman filter effectively removes noise and accurately estimates drone rotation by fusing sensor data.
• The hybrid computing approach achieved a 75% reduction in power consumption compared to traditional software-based methods.
• Demonstrated the feasibility of memtransistor-based computing for low-power drone applications.

Conclusions:

• SnS2 memtransistors offer a viable solution for energy-efficient sensor fusion in drones.
• The analog-digital hybrid computing platform significantly enhances the power efficiency of drone flight control systems.
• This technology paves the way for more capable and longer-endurance autonomous drones.