Autonomous MedRover

Autonomous MedRover is a web-connected hospital delivery rover prototype.
Users place medicine/supply delivery requests from a web portal, and an ESP32-based line-following robot navigates to the selected room route (A/B/C), performs delivery wait, and returns to base.

Project Overview

• Domain: Indoor healthcare logistics automation
• Robot: ESP32 + IR line sensors + motor driver + DC motors
• Control: PD line following with junction-based routing and recovery logic
• Cloud: Supabase (order polling + status updates)
• Frontend: React/Vite web portal for order placement and tracking

Repository Structure

• hardware_esp32/ - ESP32 firmware sketches and calibration versions
  • line_follow_v33.ino - latest route logic (A/B/C + return behavior)
  • line_follow_v32.ino - previous stable iteration
• web_portal/ - React app for ordering and order visibility
• README.md - this file

How It Works

1. User places an order in the web portal (room code A/B/C).
2. ESP32 polls Supabase for pending orders.
3. Robot updates order to in_transit and starts mission.
4. Robot follows line, counts junctions, turns based on target room.
5. At room, robot pauses (ROOM_WAIT_MS), performs U-turn, and returns.
6. On home arrival, mission completes and order is marked delivered.

Current Routing Logic (v33)

• A: turn left at junction 1, serve room, return to start.
• B: turn left at junction 2, serve room, return to start.
  • Return path is tuned to avoid unnecessary intermediate stop before home.
• C: continue on main line to top endpoint, serve room, U-turn, return to start.

Route timings and turn thresholds depend on your floor track and must be tuned in firmware constants.

Hardware Requirements

• ESP32 development board
• 5-channel IR line sensor array
• Motor driver (compatible with used pin mapping)
• 2 DC geared motors + wheels + chassis
• Battery/power module with common ground
• Black tape track on light floor

Software Requirements

• Arduino IDE (or PlatformIO) with ESP32 board support
• Required libraries:
  • WiFi.h
  • HTTPClient.h
  • ArduinoJson.h
• Node.js + npm (for web_portal)

Setup

1) Web Portal

cd web_portal
npm install
npm run dev

Create a local env file from web_portal/.env.example:

cd web_portal
cp .env.example .env

Then set:

• VITE_SUPABASE_URL
• VITE_SUPABASE_ANON_KEY

The web portal already reads these from import.meta.env.

1.1) Deploy Web Portal on Netlify

This repo includes root netlify.toml with:

• Base directory: web_portal
• Build command: npm run build
• Publish directory: dist
• SPA redirect rule to index.html

In Netlify:

1. Import the GitHub repository.
2. Build settings will auto-detect from netlify.toml.
3. Add environment variables in Site settings -> Environment variables:
   • VITE_SUPABASE_URL
   • VITE_SUPABASE_ANON_KEY
4. Trigger deploy.

2) Firmware

1. Open hardware_esp32/line_follow_v33.ino.
2. Set Wi-Fi and Supabase credentials.
3. Select ESP32 board and upload.
4. Open Serial Monitor at 115200.

Serial Commands (Firmware)

• A / B / C - choose route
• g - start mission
• s - stop
• p - print settings
• i - print IR snapshot
• w - IR watch mode
• + / - - base speed adjust
• ] / [ - Kp adjust
• > / < - Kd adjust

Key Tuning Parameters (Firmware)

• baseSpeed, turnSpeed, rightOffset
• JUNCTION_PAUSE_MS, ROOM_WAIT_MS
• BRANCH_MIN_TRAVEL_MS, LINE_END_CONFIRM_MS
• RETURN_MIN_TRAVEL_MS, RETURN_JUNCTION_COOLDOWN

Tune these values on your real track for reliable turning and home return.

Notes

• Keep secrets (Wi-Fi password, API keys) out of public repositories.
• Use calibration sketches in hardware_esp32/ to validate sensor and motor setup before full mission runs.

Future Improvements

• Dynamic map/routing instead of fixed A/B/C logic
• Obstacle detection and avoidance
• Better telemetry dashboard (battery, mission timestamps, errors)
• Safer credential handling and OTA updates