Open Source Technologies for Forest Biodiversity Monitoring
summary
A low-cost system of two open-hardware devices counts saproxylic insects with high-resolution imaging, a release mechanism and automated captures, deployed on solar + GSM connection and integrated with satellite remote sensing to produce maps and a centralized real-time dashboard. Models use YOLO and PyTorch trained on field and online images, reporting about 20–30 species and claiming ~98% accuracy in training. Hardware includes Raspberry Pi camera module, sensors and traps; the dataset remains closed despite open-source code and FAIR sharing goals.
AI-written summary · about
chapters (10)
- 0:00Introduction to project
- 1:25Open source tools used
- 2:57Field deployment and testing
- 4:26Data management and scalability
- 5:58Q&A session begins
- 6:19Accuracy of the system
- 8:09Species in the database
- 9:16Future plans and agri-tech
- 10:42Funding and business model
- 12:09Model availability and licensing
related talks
27mSmart Forestry: AI in the Service of Nature
16mKeynote: Open Source as Endurance Infrastructure for Environmental Justice
21mUsing open source software to map wildlife in the Northern Mariana Islands
19mThe Earth Data Revolution Driving Transparency in Sustainability and Security
43mPANOPTES: Open Source Exoplanet Discovery
02h 16mOffice Hours
20mEnhancing Long Term Preparedness for Forest Fires in Bhutan with Open Source
25mEmpowering weather-satellite data users with FOSS: the Pytroll case