Tom "Hutch" Ingold
OSGeo charter member, father, developer, ex-Boundless, ex-Planet, current CTO @ Earthgenome.org
Sessions
Illegal gold mining in the Amazon rainforest has escalated dramatically, resulting in deforestation, pollution, and violence against indigenous communities. This talk will explore how AI and open-source geospatial software are being used to combat this environmental crisis. We will delve into the methodologies and technologies behind Amazon Mining Watch's (AMW) efforts to monitor and mitigate illegal mining activities. By integrating machine learning models with satellite data and open-source tools, AMW provides insights that empower local activists and influence policy changes.
This session will cover the following topics:
- Overview of the environment and social impacts of illegal mining and challenges of detection
- The technological approach, beginning with purpose built machine learning models and tracing the evolution to the foundational models used today.
- Current architecture and open source components for data processing
- The use of open data, data processing strategies and information on accessing (for free!) over 100TB of processed imagery and embeddings generated to date
- Live demo of the next-gen platform
- Q/A
Links:
- https://Amazonminingwatch.org
- https://infoamazonia.org/2024/07/09/eis-o-que-a-inteligencia-artificial-achou-na-amazonia-areas-de-garimpo-dobraram/
- https://news.mongabay.com/2024/07/gold-mining-in-the-amazon-has-doubled-in-area-since-2018-ai-tool-shows/
- https://foundation.mozilla.org/en/blog/a-gold-rush-worth-stopping-ai-does-its-part
- https://pulitzercenter.org/blog/amazon-mining-watch-expands-use-ai-monitor-illegal-gold-mining
Target Audience
This talk is aimed at geospatial professionals, software developers, environmental activists, policymakers, and anyone interested in the application of AI and open-source geospatial technology for environmental protection.