Flyingcarpet: An Open Network for Building and Using Aerial Analytics Services
Duration: 00:10:35
Speaker: Julien Bouteloup
Type: Talk
Expertise: Intermediate
Event: Devcon
Date: Jun 2026
The Flyingcarpet network connects analytics-hungry businesses with a pool of data scientists who compete to create machine learning/artificial intelligence analytics-extraction models from visual data, such as drone and satellite imagery. The competition incentivisation mechanism uses bounties and a Token-Curated Registry of Opportunities (TCRO) running on the Ethereum blockchain to collect and rank machine learning model creation opportunities. From insurance companies, to agri-companies, to governments, the Flyingcarpet network enables actionable insights through rich AI-powered analytics. For example, last year, Flyingcarpet built a machine learning model that enabled a drone to autonomously count the number of coconuts in a coconut plantation in Papua New Guinea—a task which cannot be performed using satellites. The aim was to increase estimation accuracy and reduce the costs of crop yield predictions for the farmer. From a 20 minute autonomous flight, we were able to effectively collect data from the entire plantation, provide an accurate coconut count and translate that into crop yield predictions to be used on blockchain prediction platforms such as Gnosis. This information could also be used by the farmer to optimise distribution of fertilisers, water and so on. Tech paper: https://drive.google.com/open?id=10TM6bN6excBePftD6l4AoWEfZe484POp Website: https://www.flyingcarpet.network/
