Irrigation Report Automation

Automates browser-based data extraction from the IoTCrops admin portal (관수리포트 menu). Built with Playwright + Node.js. Runs a local dashboard to control everything.

---

Mac Setup (First Time Only)

1. Install Node.js

If you don't have Node.js yet, install it via Homebrew:

# Install Homebrew if you don't have it
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"

# Then install Node.js
brew install node

Verify it works:

node --version   # should be v18 or higher
npm --version

---

2. Get the Code

git clone <repo-url>
cd IrrigationReportAutomation

Or if you received the folder directly, just cd into it.

---

3. Create the .env File

The script needs login credentials. Create a file named .env in the project root:

touch .env
open -e .env   # opens in TextEdit, or use any text editor

Paste in the following (ask your supervisor for the actual values):

ADMIN_EMAIL=your_email_here
ADMIN_PASSWORD=your_password_here
ADMIN_URL=https://admin.iofarm.com/report/

Save and close. This file is gitignored and never committed.

---

4. Install Dependencies

npm install

---

5. Install the Browser (Playwright)

npx playwright install chromium

This downloads a bundled Chromium browser that the script controls. You only need to do this once.

---

6. Run It

npm start

A browser window will open, and the Dashboard will appear at http://localhost:3456 in your default browser.

---

Using the Dashboard

When the dashboard opens, configure and start the run:

Setting	What it does
Manager	Select 승진 or 진우 (or type a custom name)
Start From	Which farm to begin processing from
Mode	Normal / Watch / Learning (see below)
Max Farms	How many farms to process (use 3 for testing)

Click "Start Automation" when ready.

The dashboard shows live logs, screenshots, and pause/stop controls while it runs.

Results are saved to the data/ folder as JSON files.

---

Modes

Normal Mode (Default)

Runs fully automatically. Extracts irrigation times from charts and saves results.

Watch Mode

Same as Normal but doesn't click anything. Use this to verify the script navigates correctly without changing data.

Learning Mode

Shows the script's detection on each chart and pauses for your review:

• Green circle = detected first irrigation point
• Red circle = detected last irrigation point

If the detection is correct: press F8 to continue If the detection is wrong: click the correct points on the chart, then press F8

After 10–20 corrections the algorithm noticeably improves.

---

Daily Usage (after setup)

# Standard run
npm start

# Run for a specific manager
MANAGER="진우" npm start
MANAGER="승진" npm start

# Run with learning mode active
CHART_LEARNING="true" npm start

# Run learning mode for a specific manager
MANAGER="진우" CHART_LEARNING="true" npm start

# Analyze what the algorithm learned from your corrections
npm run analyze

---

Typical Workflow

1. npm start                          → check if data looks right

2. CHART_LEARNING="true" npm start   → if detection is off, correct it 10–20 times

3. npm run analyze                    → see what improved

4. npm start                          → now runs with better accuracy

---

Troubleshooting

Dashboard doesn't open Check the terminal for a line like Dashboard server started at http://localhost:XXXX. If port 3456 is taken, the server picks the next available port automatically.

Script hangs and does nothing You're probably in Learning Mode and need to press F8 to continue.

"Cannot find module" or import errors Run npm install again — a dependency is missing.

Browser opens but login fails Double-check your .env file has the correct email, password, and URL with no extra spaces.

Kill all Node processes if something is stuck

killall node

---

Project Structure

IrrigationReportAutomation/
├── irrigation-playwright.js   # Main automation script
├── src/                       # Core modules (auth, navigation, chart analysis)
├── training/                  # Saved learning data (gitignored)
├── data/                      # Extracted results (gitignored)
├── .env                       # Credentials (gitignored, you create this)
├── package.json
└── README.md

---

Tech Stack

• Playwright — browser automation
• Node.js (ES Modules) — runtime
• TensorFlow.js — TCN model for irrigation detection
• HSSP Algorithm — baseline chart analysis (Highest Slope Start Point)
• Dashboard — HTTP server with Server-Sent Events for real-time UI