AgroFleet: Community Disaster & Crop Resilience Network

Inspiration

Extreme weather events and localized agricultural crises severely impact marginal farming communities. Smallholder farmers often lack real-time coordination tools for disaster recovery, shared machinery management, and rapid crop threat mitigation. AgroFleet was inspired by the need for a resilient, community-first AI fleet system that runs in the background to handle heavy operational logistics without technical overhead for farmers.

What it does

AgroFleet is an autonomous multi-agent fleet network built under the Good Neighbor track. It coordinates agricultural resources, predicts local crop risks, and manages crisis responses automatically:

• Disaster Response Agent: Monitors hyper-local weather alerts and automatically coordinates emergency machinery or harvesting schedules before severe weather hits.
• Community Resource Agent: Manages fair peer-to-peer sharing of autonomous tractors and drones across neighboring farms.
• Crop Health Agent: Runs diagnostic checks and alerts neighboring farms of pest outbreaks or disease vectors.

How we built it

• AWS Strands Agents SDK: Wraps core agent interactions, handling task delegation, state management, and inter-agent communication protocols.
• Gemini AI Backend: Powers high-level inference, spatial reasoning, and complex decision-making for agricultural risk assessment.
• Google Cloud Infrastructure: Serves as the high-availability execution context and data ingestion layer for IoT telemetry.

Challenges we ran into

Integrating a hybrid cloud architecture where AWS Strands Agents interface seamlessly with GCP-hosted Gemini inference required custom async adapters to maintain real-time response limits for edge alerts.

Accomplishments that we're proud of

Successfully building a true Multi-Agent System (MAS) that bridges two distinct cloud and AI frameworks (AWS and GCP) to deliver zero-friction coordination for vulnerable rural communities.

What we learned

Optimizing agent communication loops using AWS Strands SDK significantly reduces latency when coordinating multi-step automated tasks in low-bandwidth agricultural regions.

What's next for AgroFleet

• Integration with regional meteorological APIs for automated insurance dispatch.
• Deploying low-power edge agents directly onto physical agricultural hardware.

Built With

• agriculture
• ai-agents
• aws-strands-sdk
• gemini-api
• google-cloud
• multi-agent-system
• python