Neuromorphic Sensor Integration for Plant Health Monitoring

alternate text

Supervision:
INI - Dr. Chiara De Luca (contact person)
Prof. Giacomo Indiveri

FHNW: Prof. Gerd Simons

This project will be carried out in collaboration with the FHNW Institute for Sensors and Electronics[1]

 

Motivation:
Monitoring plant health is crucial for early detection of pests, identifying anomalies, and ensuring timely interventions. While numerous sensors are available for this purpose, selecting the most effective ones and eliminating redundancy remains a challenge. Additionally, transmitting large volumes of data to the cloud is power-intensive, especially in resource-constrained environments. To address these challenges, local preprocessing is essential to reduce data load and enhance efficiency. Leveraging neuromorphic hardware provides a promising approach to achieve low-power, real-time processing for plant status monitoring.

 

Project:
The goal of this project is to implement Spiking Neural Networks (SNNs) on mixed-signal DYNAPSE chips to integrate sensor data for plant health monitoring. The student will:

  • - Identify essential sensors and determine which are redundant.

  • - Design and implement an SNN-based approach for local data preprocessing.

  • - Implement the network on the mixed-signal DYNAP-SE chip[2]

  • - Use data and experimental setups provided in collaboration with the lab ### to validate the system's performance.

This interdisciplinary project merges cutting-edge neuromorphic computing with real-world agricultural applications, providing an exciting opportunity for impactful research.

 

Challenges and Opportunities:

  • - Hands-on experience with real-world neuromorphic hardware, including understanding its constraints and advantages.

  • - Modeling complex biological systems such as plant health from real-world data, requiring innovative approaches.

  • - Opportunity to contribute to a sustainable solution for agriculture by addressing power-efficient data processing.

 

Student Expectations and Opportunities:
The ideal candidate for this project should have:

  • - A strong foundation in Python programming and experience with Git version control.

  • - Basic knowledge of Spiking Neural Networks (SNNs) and enthusiasm for neuromorphic computing.

  • - A background in computational mathematics, physics, biology, or a related field (preferred).

The student will have access to neuromorphic chips at INI and datasets provided by ###, offering a unique hands-on learning experience. This project is an excellent fit for students passionate about interdisciplinary work at the intersection of physics, biology, and computer science.

 

References:

[1] https://www.fhnw.ch/en/research-and-services/engineering/laboratories-of-the-fhnw-school-of-engineering/innovation-garden

[2] Moradi, S., Qiao, N., Stefanini, F., & Indiveri, G. (2017). A scalable multicore architecture with heterogeneous memory structures for dynamic neuromorphic asynchronous processors (DYNAPs). IEEE transactions on biomedical circuits and systems, 12(1), 106-122.

back
back