STATION SAFETY SYSTEM

SafetyIQ

A worldwide, open-source ham radio station safety monitor — built to protect antenna towers when no one is home.

Introduction

SafetyIQ is a worldwide ham radio station safety monitoring system. It runs 24/7/365 headless on a Raspberry Pi or any small PC, continuously watching real-time weather data from dozens of global sources, and automatically warns the operator when wind speeds are approaching the structural limits of their antenna towers.

When a threat is detected, the operator receives an alert via Email and Telegram — containing the current threat level, gust reading, an AI-generated narrative, and a one-tap secure link to the remote control panel. From that panel, towers can be lowered, alerts acknowledged, or all-clear declared from anywhere in the world.

If the operator does not respond within a configurable time window, SafetyIQ acts autonomously — lowering towers, routing antennas to a dummy load — protecting the station from storm damage without any human intervention.

Why SafetyIQ exists

Ham radio operators invest thousands of euros in antenna towers and equipment. A single storm while the operator is away from home — unable to lower the antennas in time — can destroy years of work in minutes. No commercial solution exists for this specific problem.

SafetyIQ fills that gap. It is not a weather app. It is not a logging tool. It is a dedicated, always-on safety watchman for the station — monitoring continuously, alerting intelligently, and acting when the operator cannot.

How it works

SafetyIQ pulls real-time and forecast weather data from over 40 sources simultaneously — METAR from the nearest airports, Open-Meteo global forecast and ERA5 historical data, NOAA ISD surface observations, Blitzortung real-time lightning, NHC and JTWC tropical cyclone tracks, and regional meteorological agencies from 26 countries including IPMA, DWD, BOM, NWS, JMA, METno, SMHI, and more.

An AI engine (Claude claude-sonnet-4-6 by Anthropic) assesses the compound threat across all data simultaneously, generates a human-readable operator narrative, and drives a five-state threat tracker — from WATCHING through DEVELOPING, IMMINENT, PASSING to CLEAR. A parallel rule-based engine runs every 10 seconds without any AI key, ensuring the system always works regardless of connectivity or account status.

When a threat crosses configurable thresholds, the alert chain begins. If the operator responds via the remote panel, the system acknowledges and continues monitoring. If no response arrives within the countdown window, SafetyIQ executes the safety actions and sends a follow-up notification: “SafetyIQ acted automatically: tower lowered.”

Key features

• Worldwide — configure with a single lat/lon coordinate; SafetyIQ adapts automatically to any location on Earth
• Headless operation — runs without a browser, without a screen, without any user interaction
• AI threat assessment — compound pattern detection, human-readable narrative, trajectory memory
• Five-state threat tracker — WATCHING → DEVELOPING → IMMINENT → PASSING → CLEAR
• Three-layer presence detection — manual toggle, Thetis TCI radio detection, phone on LAN
• Alert delivery — Email and Telegram with secure one-tap remote link
• Autonomous safety actions — lowers towers, routes to dummy load when operator is unreachable
• Map-first dashboard — full-screen animated weather map with overlay threat indicators
• Progressive Web App — installable on phone or tablet home screen, works offline
• Pattern learning — logs every event to a local database, builds location-specific threat profiles over time
• Open-source — free, no subscription, no cloud dependency

What SafetyIQ is not

SafetyIQ is a watchman, not a station controller. It never inhibits radio operation. Wind approaching the structural threshold does not stop the operator from transmitting — it triggers an advisory. The operator always decides. SafetyIQ only acts autonomously when the operator is confirmed absent and the countdown has expired.

Status

SafetyIQ is in active development, running live at the CT2IRY station in Lisbon, Portugal as the primary testbed. The GitHub repository is currently private during the rapid development phase. When the system reaches a stable public release it will be open-sourced on GitHub and a full download and configuration guide published here.