AeroMind – Agentic AI Autonomous Campus Delivery Drone

Implementation and simulation results of the developed project.

Inspiration

Campus delivery is often inefficient, slow, and manpower-dependent. We were inspired by the growing need for contactless, fast, and autonomous delivery systems in large campus environments such as universities and tech parks. Advances in agentic AI, computer vision, and UAV technology motivated us to build a system where drones can think, plan, and act independently instead of following rigid scripts.

What it does

AeroMind allows users to request deliveries through a web platform. An agentic AI system intelligently assigns a drone, plans the route, avoids obstacles in real time, and ensures secure delivery using OTP or face recognition. Admins can monitor drone status, location, and delivery progress live through a dashboard.

How we built it

We designed a full-stack system integrating hardware and software: A web interface (MERN stack) for delivery requests and admin monitoring Agentic AI modules for decision-making, task allocation, and replanning Autonomous navigation using SLAM and depth-based obstacle detection Onboard compute using Raspberry Pi / Jetson for vision and control ROS + MAVSDK for drone communication and flight control The system continuously senses the environment and adapts its actions dynamically.

Challenges we ran into

Integrating real-time vision processing with drone flight control Ensuring safe navigation in dynamic campus environments Synchronizing AI decision-making with physical hardware constraints Managing latency between the web system and drone telemetry

Accomplishments that we're proud of

Successfully designed an agentic AI–driven delivery workflow Built a scalable web-based drone management system Implemented real-time obstacle detection and autonomous navigation Designed a secure delivery verification mechanism Created a system that can be extended to real-world smart campuses

What we learned

Practical application of agentic AI in robotics Integrating computer vision, ROS, and web systems Designing fault-tolerant autonomous systems Bridging the gap between AI logic and real-world hardware execution

What's next for AeroMind – Agentic AI Autonomous Campus Delivery Drone

Multi-drone swarm coordination Weather-aware route planning Battery-aware task optimization Integration with campus ERP systems Full real-world deployment and testing