Related Experiment Video
Updated: Jun 19, 2026
09:13
A Fully Automated and Highly Versatile System for Testing Multi-cognitive Functions and Recording Neuronal Activities in Rodents
Published on: May 3, 2012
Self-calibrating neuromorphic system for adaptive environmental sensing
Anantharaman Prasad1, S Sofana Reka2, Prakash Venugopal2
1School of Electronics Engineering, Vellore Institute of Technology, Chennai, India.
Frontiers in Artificial Intelligence
|June 18, 2026
Summary
This study introduces a novel self-calibrating neuromorphic system for adaptive soil moisture sensing. It uses Spiking Neural Networks (SNN) on microcontrollers to reduce energy use and improve accuracy in precision agriculture.
Area of Science:
- Neuromorphic Computing
- Internet of Things (IoT)
- Precision Agriculture
Background:
- Conventional soil moisture sensors suffer from drift, high energy use, and poor adaptability.
- These issues lead to inefficient irrigation and unreliable data, especially in remote areas.
- Frequent recalibration is often impractical for existing sensor technologies.
Purpose of the Study:
- To develop a novel self-calibrating neuromorphic system for adaptive soil moisture sensing.
- To address the limitations of conventional sensors in precision agriculture.
- To enable accurate, real-time environmental monitoring with reduced energy consumption.
Main Methods:
- Leveraged Spiking Neural Networks (SNN) deployed on a low-power STM32H563ZI microcontroller.
- Implemented autonomous sensor recalibration to mitigate drift.
- Utilized event-driven computation for reduced energy consumption.
Main Results:
- Achieved a Mean Absolute Error (MAE) of 0.4557 and Root Mean Squared Error (RMSE) of 0.5850.
- Reduced baseline sensor drift from 5.3% to 1.6% over two months.
- Outperformed Isolation Forests and Autoencoders in predictive accuracy for soil moisture sensing.
Conclusions:
- The proposed SNN-based system offers a scalable, low-power solution for precision agriculture and environmental monitoring.
- Demonstrated effective deployment of neuromorphic learning on constrained microcontroller hardware.
- Opens new avenues for resilient, decentralized intelligence in IoT applications.
