Sky Guardians: Revolutionizing Wildlife Monitoring with Drones and IoT
summary
Claims a working system that automates 24/7 detection, localization and active scaring of geese, built from cheap Raspberry Pi audio sensors with custom PCBs and solar power, a spectrogram ML model (detection threshold 0.7), and orchestration on AWS Lambda + S3 + EC2 auto-scaling. Over 100 sensors run in the field, each producing ~6–700 MB/day, feeding alerts that cue drones to chase flocks; the method reduces damage (national costs > €55M cited) and becomes more effective over months rather than losing efficacy.
AI-written summary · about
chapters (14)
- 0:00Introduction to wildlife monitoring
- 2:52Problem of goose damage in agriculture
- 6:55Drones as goose deterrents
- 10:44Developing audio sensors for geese detection
- 14:59Hardware challenges and solutions
- 17:56Powering sensors with solar panels
- 20:29Combining software and hardware development
- 24:27Shifting focus to sensors over drones
- 27:20Audio recognition with machine learning
- 31:34Running models on Raspberry Pi and AWS
- 34:52Data processing and dashboard
- 37:29Future of automated drone systems
- 39:37Ecological impact and population control
- 40:06Q&A session
related talks
15mThe Internet of Wild Things: Acoustics, AI and Conservation
30mProtecting Wildlife with LoRaWAN
01h 00mBeeIoT - Enter the World of Bees
24mWildlife Protection
09mAgri Drones Programmed with Java
39mFrom IoT to Human interactions
41mLeveraging Next Generation Cellular Networks for Drone Telemetry and Paylo...
47mVisualizing Aircraft Data: Oracle IoT Cloud Service, Raspberry Pi, and Radio Surveillance