A curated, community-maintained hub for edge computing and edge AI methods, tools, datasets, and deployments in conservation.
This survey is intended to consolidate projects and resources for deploying edge AI for conservation. This page is maintained by Jenna Kline. Contributions are welcome! Please submit additions or corrections as Github pull requests.
Edge AI is transforming wildlife monitoring and ecosystem science by running perception, decision-making, and data analysis on-device and in the field . Relevant projects, tools, and papers, span multiple subfields (computer systems, AI, robotics, acoustics, ecology). This repo aggregates references and resources: what to use, how to deploy, and where to learn from real-world efforts. This repo is intended to be a living document that evolves with the field.
This project is inspired by excellent modality-specific surveys, including Everything I know about ML and camera traps maintained by Dan Morris and Computer Vision and Aerial Imagery for Wildlife Conservation maintained by Dan Morris and Benjamin Kellenberger. This repo aims to broaden to multi-modal edge AI for conservation with a consistent taxonomy, runnable examples, and deployment-first guidance.
This project prioritizes:
Out of scope: purely cloud-only analytics without an edge component.
Wild Edge (WildLabs Group) Community for edge AI in conservation wildlabs.net
Microsoft Project SPARROW Solar-Powered Acoustic and Remote Recording Observation Watch; Biodiversity edge‑AI initiative/platform; Solar-powered sensors collect biodiversity data (camera traps, acoustics), processed on low-energy edge GPUs with PyTorch wildlife AI models, information transferred to cloud via low-Earth orbit satellites Announcement
SmartWilds Edge AI for animal land use patterns, species ID; multi-modal sensors including camera traps, bioacoustics, drones, and GPS trackers Project Page
Robinson Crusoe Island Project Delivering edge computing on Robinson Crusoe Island (Chile) Article
NVIDIA Jetson (Nano / Orin Nano) Embedded GPU modules for on-device AI Jetson Nano, Orin Nano
Raspberry Pi Single-board computers for far‑edge deployments raspberrypi.com
Google Coral Edge TPU accelerators and dev boards coral.ai, Docs
Seeed Studio LoRaWAN Dev Kit LoRaWAN gateway/dev kit for LPWAN edge sensing Product
Edge Impulse Edge ML platform company, specializing in data→model→deploy; focused on industry applications but usable for conservation, see Blog on Adaptive Camera Trap with GPT-4o edgeimpulse.com – Intro to Edge AI Course
ICICLE AI Institute US NSF funded AI institute for Intelligent Cyberinfrastructure with Computational Learning in the Environment (ICICLE); focus on democratizing AI for digital agriculture and animal ecology applications ICICLE Website – ICICLE GitHub Repo
SAGE Grande Testbed The Sage Grande Testbed (SGT) is building a cutting-edge artificial intelligence (AI) cyberinfrastructure to support advanced AI research. Funded by the NSF Office of Advanced Cyberinfrastructure, provides access to AI-enabled edge computing resources and software tools integrated with sensors—including infrared and RGB cameras, microphones, and a variety of atmospheric and air quality instruments—deployed across natural, urban, and wildfire-prone environments, with networking capabilities that support real-time hazard reporting. SAGE Website
BirdWeather AI powered bioacoustics platform https://www.birdweather.com/
WildWing DIY open-source drone; developed for autonomous animal behavior monitoring missions. WildWing Project
Sentinel Hub Smart camera trap plaform from non-for profit ConservationXLabs; aimed at democratizing AI for wildlife and disease monitoring. Sentinel Project
Mini AI Wildlife Monitor DIY AI edge compute based wildlife detection with Raspberry Pi AI Camera. See Discussion post on WildLabs and YouTube tutorial
Animl Camera trap platform - see ‘Lessons learned from deploying and managing wireless camera trap networks in remote environments’ for edge AI specific guidelines – Animal Camera Guide, YouTube video, and Article.
BirdNET-Pi A realtime acoustic bird classification system for the Raspberry Pi 5, 4B 3B+ 0W2 and more https://www.birdweather.com/birdnetpi
YOLO Small, real‑time models for detection and classification, ‘You Only Look Once’ https://docs.ultralytics.com/
MobileSAM Segment Anything Model optimized for edge https://github.com/ChaoningZhang/MobileSAM
WL-YOLO Designed for lightweight wildlife for real-time detection in complex forest environments. Paper
Edge computing in wildlife behavior and ecology — Trends in Ecology & Evolution 39(2):128–130, 2024. Yu, H., et al. Paper
Characterizing and Modeling AI-Driven Animal Ecology Studies at the Edge — 2024 IEEE/ACM Symposium on Edge Computing (SEC), Rome, Italy. J. Kline, A. O’Quinn, T. Berger‑Wolf, C. Stewart. Paper – GitHub Repo
Environment-Aware Dynamic Pruning for Pipelined Edge Inference — arXiv preprint (2025). O’Quinn, Snedeker, Zhang, Kline. Paper
ML Field Planner: Analyzing and Optimizing ML Pipelines for Field Research — PEARC ’25: Practice and Experience in Advanced Research Computing, 2025. Stubbs, J., Balasubramaniam, S., Khuvis, S., Withana, S., Vallabhajosyula, M. S., Cardone, R., Garcia, C., Freeman, N., Guzman, C., Plale, B., Ramnath, R., & Berger-Wolf, T. Paper
Broad-scale applications of the Raspberry Pi: A review and guide for biologists — Methods in Ecology and Evolution, 2021. Jolles, J. W. Paper
acoupi: An Open-Source Python Framework for Deploying Bioacoustic AI Models on Edge Devices - arXiv preprint arXiv:2501.17841., Vuilliomenet, A., Balvanera, S. M., Mac Aodha, O., Jones, K. E., & Wilson, D. Paper Repo Docs
Wildlife Monitoring on the Edge: A Performance Evaluation of Embedded Neural Networks on Microcontrollers for Animal Behavior Classification — Sensors 21(9):2975, 2021. Dominguez‑Morales, J. P., Duran‑Lopez, L., Gutierrez‑Galan, D., Rios‑Navarro, A., Linares‑Barranco, A., & Jimenez‑Fernandez, A. Paper
Building of an edge-enabled drone network ecosystem for bird species identification — Ecological Informatics 2022. Das, N., et al. Paper
WildWing: An open-source, autonomous and affordable UAS for animal behaviour video monitoring — Methods in Ecology and Evolution, 2025. Kline, J., Zhong, A., Irizarry, K., Stewart, C. V., Stewart, C., Rubenstein, D. I., & Berger‑Wolf, T. Project Page Paper
WildLive: Near Real-time Visual Wildlife Tracking onboard UAVs - CV4Animals Workshop at CVPR 2025. Dat NN, Richardson T, Watson M, Meier K, Kline J, Reid S, Maalouf G, Hine D, Mirmehdi M, Burghardt T. Paper
Semi-supervised Visual Tracking of Marine Animals Using Autonomous Underwater Vehicles - Int J Comput Vis 131, 2023. Cai, L., McGuire, N.E., Hanlon, R. et al. Paper
AI-Driven Real-Time Monitoring of Ground-Nesting Birds: A Case Study on Curlew Detection Using YOLOv10 — arXiv preprint arXiv:2411.15263. Chalmers, C., Fergus, P., Wich, S., Longmore, S. N., Walsh, N. D., Oliver, L., Warrington, J., Quinlan, J., & Appleby, K. Paper
An IoT System Using Deep Learning to Classify Camera Trap Images on the Edge — Computers, 2022. Zualkernan, I., Dhou, S., Judas, J., Sajun, A. R., Gomez, B. R., & Alhaj Hussain, L. Paper
Energy-Efficient Audio Processing at the Edge for Biologging Applications — Journal of Low Power Electronics and Applications, 2023. Miquel, J., Latorre, L., & Chamaillé‑Jammes, S. Paper
Real-time alerts from AI-enabled camera traps using the Iridium satellite network: A case-study in Gabon, Central Africa — Methods in Ecology and Evolution, 2023. Whytock, R. C., et al. Paper
Reliable and efficient integration of AI into camera traps for smart wildlife monitoring (with on-device continual learning) — Ecological Informatics, 2024. Velasco‑Montero, D., Fernández‑Berni, J., Carmona‑Galán, R., Sanglas, A., & Palomares, F. Paper
Towards a standardized framework for AI-assisted, image-based monitoring of nocturnal insects — Philosophical Transactions of the Royal Society B , 2024. Roy, D. B., Alison, J., August, T. A., et al. Paper
Towards scalable insect monitoring: Ultra-lightweight CNNs as on-device triggers for insect camera traps — Methods in Ecology and Evolution, 2025. Gardiner, R. J., Rowlands, S., & Simmons, B. I. Paper
Assessment of technological developments for camera-traps: a wireless network and advanced image recognition — Wildlife Society Bulletin, 2022. Meek, P., et al. Paper
A computer vision enhanced IoT system for koala monitoring and recognition - Internet of Things, 2025. Trevathan, J., Tan, W.L., Xing, W., Holzner, D., Kerlin, D., Zhou, J. and Castley, G., Paper
We welcome issues and PRs for:
MIT — see LICENSE.