Aerodava

Inspiration

Aerodava started in February 2026 at HackFusion 2026, where we tackled healthcare accessibility in rural and hard-to-reach regions with a drone-based medical delivery idea. We didn't qualify for the next round, but the idea stuck with me.

I kept building it solo for an international ideathon by CareerPrep Tech, where it placed 15th. That's when the cracks in a drone-first approach became obvious — cost, maintenance, theft, weather, and scale.

Then I found this hackathon Hack2Heal , 3 days before the submission deadline and picked up aerodava to improve and make a submission For Hack2Heal, the question became bigger than delivery: How can a patient stay connected to healthcare when access itself becomes difficult?

That reframing is what turned Aerodava into a full healthcare-access platform, with drones kept only as an emergency layer.

What it does

Aerodava connects patients to the right healthcare resource at the right stage:

• AI Medical Copilot — explains reports/prescriptions in simple language, and answers symptom questions with conservative guidance + red-flag escalation
• Doctor Connect — patients share info with doctors; AI summarizes, doctors decide
• Patient Health Portfolio — history, reports, meds, and prescriptions in one place
• Prescription Understanding & Medicine Access — uploads get explained, or extracted straight into a pharmacy/cart flow
• Human Mesh — nearby-device relay for alerts when internet is down
• Emergency Drone Logistics — drones only for disaster/no-access scenarios, not routine delivery

(Full breakdown in the GitHub repo.)

How we built it

• Frontend: Next.js 14 (App Router), React 18, TypeScript
• UI: Tailwind CSS + shadcn/ui + Lucide icons, with a custom healthcare palette
• Data/AI: Client-side mock architecture — seeded data + async service layer + a simulated AI service, built so it can be swapped for real APIs/LLMs later without a rebuild

Challenges we ran into

• Realizing drones as a daily-use solution had real financial/operational risk — damage, theft, maintenance — so we restricted them to emergency-only use
• Deciding how much to trust AI in healthcare — pivoted from "AI handles everything" to AI handling basic cases and red-flagging/escalating serious ones to a human doctor layer
• Understanding rural user needs beyond "no access" — surfaced smaller but real pain points like not understanding prescriptions
• Fitting all of this into a hackathon timeframe without faking integrations that don't exist yet

Accomplishments that we're proud of

• Evolving Aerodava from a single drone idea into a layered healthcare platform without losing the original concept
• Placing 15th at the CareerPrep Tech international ideathon while developing this solo
• Getting a working end-to-end prototype: Patient → AI Understanding → Risk Escalation → Doctor → Prescription → Medicine Access
• A clean, honest split of responsibility: > AI handles the routine. Humans handle the clinical. Hardware handles the unreachable.

What we learned

• Thinking at the ground level and mapping edge cases matters more than chasing a technically flashy idea — that's exactly what killed the drone-only model
• Healthcare pain points often live in the small gaps (prescriptions, communication, escalation) — not just the big obvious ones
• Edge cases should be designed for from day one, not patched in later

What's next for Aerodava

• Talk to real hospitals, doctors, and rural users to validate (or break) our assumptions
• Rethink the drone system for range, connectivity, and real deployment cost
• Move from simulated AI to real, possibly domain-specific/fine-tuned models
• Build out real pharmacy, hospital, and data-consent integrations
• Prioritize finishing and hardening the prototype into something market-ready, focusing on niches and edge cases rather than new features

Built With

• eslint
• github
• lucide-react
• next.js
• node.js
• npm
• react-native
• shadcn-ui
• tailwind-css
• typescript
• vercel