Automate Singapore COE Price Analysis & Purchase Timing with GLM-4.5 AI Predictions

Introduction

Automates Singapore COE price tracking, predicts trends using AI, and recommends optimal car purchase timing. Scrapes LTA data biweekly, analyzes historical trends, forecasts next 6 bidding rounds, and sends alerts when buying windows appear—saving time and identifying cost-saving opportunities.

How it Works

Biweekly trigger scrapes LTA COE data → processes historical trends → AI predicts 6-month prices → compares current vs forecast → generates buy/wait recommendations → alerts sent via Gmail or Telegram.

Setup Steps

1. Add NVIDIA/OpenAI API credentials in n8n
2. Connect Google Sheets for data storage
3. Authenticate Gmail/Telegram for notifications
4. Schedule trigger for Wednesdays 8PM SGT
5. Configure alert thresholds in conditional nodes

Workflow

Schedule Trigger → HTTP Request (Scrape LTA) → Data Processing → Google Sheets (Store) → AI Prediction → Analysis Engine → Conditional Logic → Gmail/Telegram Notification

Workflow Steps

1. Scraping: Extract COE prices from OneMotoring
2. Processing: Calculate moving averages, volatility, seasonal trends
3. Storage: Save to Google Sheets with timestamps
4. Prediction: AI forecasts next 6 bidding rounds
5. Analysis: Compare current vs predicted prices, generate recommendation
6. Notification: Alerts via email/Telegram

Prerequisites

NVIDIA/OpenAI API key, Google account (Sheets), Gmail/Telegram for notifications, basic COE category knowledge

Use Cases

First-time buyers monitoring price dips, fleet managers timing bulk purchases

Customization

Add economic indicators, integrate car loan calculators, track parallel imported car prices

Benefits

Saves hours of manual monitoring, captures 10–15% price dips, provides data-driven purchase timing (potential $5K–$15K savings)