ApiSense: A Weather-Aware Multimodal System to Detect Environmental Pesticide Contamination via Honeybees as Biosensors
CSEF · 2026 Zoology (Senior Division)
Overview
Pesticide contamination poses a major threat to the ecosystem, costing $10-15 billion in environmental and health costs. Traditional monitoring methods such as soil and water sampling cannot effectively cover a large area or provide early warnings of contamination. However, honeybees forage over 7 km2 daily and are sensitive to extremely low concentrations of pesticides. While researchers have shown that pesticide exposure changes bee foraging behavior, few studies use foraging behavior to detect pesticide exposure. Acoustic-based approaches use shallow classifiers that cannot capture how acoustic signals change over time, while most studies ignore the impacts of weather on colony behavior. I present ApiSense: a monitoring system that uses bees as biosensors to detect pesticide contamination. ApiSense uses a gating mechanism to separate toxicological responses from those due to weather, transformers to capture long-term temporal dependencies, and a bidirectional cross-attention module to fuse foraging and acoustic signals. I applied the pesticide bifenthrin via outside feeders to 16 beehives over 15 days and collected 20,160 samples. ApiSense achieves 80.2% accuracy within 3 days of exposure, and 79.6% mean accuracy, outperforming all baseline models by at least 8%. Acoustic stability and navigation precision were identified as early indicators of pesticide exposure, aligning with toxicological findings that pesticides impair motor control and spatial memory before colony decline. ApiSense can benefit beekeepers, farmers, and regulatory agencies. At just $120 per hive, ApiSense can be deployed across the existing 2.7 million U.S. beehives to build a pesticide monitoring network that cannot be achievable by traditional methods.
Competition history
- CSEF 2026
Related projects
ISEF · 2026
ApiSense: A Weather-Aware Multimodal Framework for Environmental Pesticide Contamination Detection Using Honey Bees as Biosensors
ISEF · 2023
BeeWell: Developing an AI-Based Bee Health Assessment System Utilizing Computer Vision and Acoustic Signal Processing
ISEF · 2024
Sensor-Assisted, Noninvasive Monitoring for Operational, Investigative and Predictive Hive Health Management for Honeybee Colonies – Sustainable Bee Keeping
CSEF · 2023
BeeWell: Developing an AI-based Bee Health Assessment System Utilizing Computer Vision and Acoustic Signal Processing
ISEF · 2024
Year 5: Engineering a Novel IoT Trapping System With a Multiple Linear Regression Model Towards Eco-Friendly, High-Efficacy, Low-Cost Honey Bee Pest Treatment and Management
ISEF · 2022
Secret Sounds of Bees: Analysis of Honey Bee Vibroacoustics Using Hidden Markov Models
CSEF · 2026
OxiGuard AI: A Machine Learning Model for Predicting Oxalic Acid Residue to Optimize Varroa Mite Control in Honeybees
CSEF · 2026
BeeCast: A Novel Predictive Model Using Neural Networks to Predict Colony Collapse
Closest projects by meaning, across every fair and year in the corpus.
Browse more like this
Source: California Science & Engineering Fair public projects