Key words: Oyster Behavior, Hall Effect Sensor, Deep Learning, Real-time Monitoring, ESP32, AWS, IoT, FFT analysis
This study presents a real-time oyster gaping behavior monitoring system based on FFT and machine learning models, providing insights into spawning, feeding, and aging processes. The system uses ESP32 and AWS to enable high-frequency data collection and cloud storage. Experimental results demonstrate the system’s effectiveness in detecting critical physiological activities in oysters. This technology holds promising applications in aquaculture and ecological monitoring.
Analyzing oyster gaping behavior is essential for understanding oyster health, behavior, and environmental interactions. This study builds on current methods to develop a fully automated, real-time system for measuring oyster gape, with the goal of gaining insights into key physiological conditions such as spawning, feeding, and aging.
Our system combines HAL 2425 Hall Effect sensors paired with magnets and an ESP32 microcontroller to accurately measure the distance between each sensor and magnet attached to an oyster shell. Data is collected at 10 Hz and transmitted via Wi-Fi or cellular network to AWS cloud storage, supporting remote analysis and continuous monitoring. Integrating cloud computing into this setup reduces the need for constant laboratory presence by allowing remote access.
To ensure accuracy, each sensor undergoes a linearization process before deployment, converting ADC values into precise distance measurements that reflect oyster gape. This continuous data stream can then be processed through FFT to analyze power spectral density. Spectral analysis reveals notable power spikes in the 0.3–1.3 Hz range, signaling spawning events. Initial results suggest that applying a 0.1 dB threshold within this range provides reliable spawning detection. This dataset also supports machine learning models for predicting spawning, feeding, and other behavioral patterns.
This research introduces a scalable and innovative solution for real-time oyster gape monitoring, with applications in aquaculture management, ecological research, and marine biology. By combining sensor technology, microcontrollers, cloud computing, and AI, this system offers a solid foundation for advancing bivalve behavioral studies.
Oysters used in this research were sourced from the Mississippi Gulf, provided by the Gulf Coast Research Laboratory (GCRL), and monitored under controlled lab conditions optimized for growth and behavior observation. Data from this study will be made available through the MBRACE project data repository.