The ECAAS team participated in two conferences in March, the ICTforAg virtual conference and the World AgriTech summit held in San Francisco, CA. The conferences allowed the team to network with experts in the agriculture and technology ecosystem and expand the community committed to capturing and sharing ground-truth training datasets for agricultural applications.
ICT4Ag is a free virtual conference that spanned two days and covered all of the world’s time zones. The conference themes explored the current and future trends in information communication technologies and agriculture, and their intersection with locally-led development, climate, digital inclusion, and digital data sovereignty. ECAAS’ session specifically focused on how we are delivering the promise of advanced analytics for smallholder farmers.
In low and middle-income countries, 65 percent of working adults rely on the agricultural value chain for their livelihoods. Agriculture is inherently risky, often reliant on rainfed irrigation, and subject to extreme weather events due to climate change. Emerging technologies promise improvements like making farming easier, de-risking investments, and improving working conditions. However, large gaps and silos exist in the data ecosystem needed to build technologies and develop informed agricultural policies.
ECAAS is working to bridge those silos and scale the collection and sharing of ground-truth training data needed to build applications that support smallholder farmers worldwide. This includes supporting crop area estimations, yield forecasting models, and field boundary detection data that can feed into machine learning and artificial intelligence applications. Making data accessible and easy to produce at a low cost requires innovations that can be replicable and scalable. ECAAS specifically supports innovations pushing the boundaries in ground truth data collection through our Acceleration Facility.
During the presentation at ICTforAg, we enthusiastically welcomed participation and feedback on where the community thinks the crop analytics sector should go next. Participants shared that they want to see actionable and localized solutions in applications. Participants recognized machine learning as key to helping the community reach the goal of providing local data for local farmers, but the community wanted to see greater emphasis placed on sharing ground-truth data and making it openly sourced. Lastly, the community wanted to see initiatives and applications that have been tested and will stick around for the long term instead of learning and relying on new pilot applications that might not be permanent solutions. The future of agriculture is accessible and actionable data for smallholder farmers, and ECAAS will continue to center this feedback in future phases of our work.
The amazingly talented Tanya Gadsby with the Fuselight Creative graphically recorded notes from the session. The photo below summarizes important takeaways from the small group discussions around best practices for collecting and sharing agricultural data.