Why I built it
It started with running.
I built Whether To Run, a small app that answered one question: is it safe to go out right now? Heat, humidity, air quality, sun load. Not a color-coded icon. An actual heat stress model.
Building it meant going deep on the physics. WBGT. The Liljegren model. NIOSH work/rest ratios. TB MED 507. And somewhere in that work I realized the science was never the hard part. Heat stress thresholds have been settled for forty years. The hard part is getting the number to the person who has to act on it, early enough for the number to matter.
A runner can check an app before leaving the house. A crew of 300 pouring concrete in Orlando cannot. Nobody is checking a phone. The decision has to arrive on its own.
That pointed me at job sites. Once I got in front of safety managers, the same thing came up in every conversation: heat is a slow problem, lightning is a fast one. Heat gives you a shift to plan around. Lightning gives you minutes, and the standard practice is still a stopwatch, a phone tree, and somebody's read of the sky. Miss it and you have a fatality. Call it wrong the other way and you just sent 300 people to their trucks for nothing.
So I started with lightning.
What it does
Auger watches for cloud-to-ground lightning inside a radius you set around the site. When a strike lands inside it, the stand-down goes out as a text message to every phone on the crew list. Every new strike resets a 30 minute clock. When the clock runs out, the all clear goes out the same way.
Heat, wind, and air quality run against your thresholds on the site dashboard.
Every condition, alert, and decision gets written to a Decision Log you can export. What the site saw, when it saw it, who got told, and what happened next.
No hardware. No sensors to mount. Nothing for the crew to download. It's a text message.
What I'm actually building
Detection is a commodity. You can rent lightning data from four different vendors before lunch.
The unsolved problem is orchestration. Moving hundreds of people to shelter reliably, then being able to show six months later exactly what you knew and when you knew it. That is what an incident investigation asks for and it is what almost nobody can produce.
So Auger is a decision engine, not a weather dashboard. It doesn't hand you radar and leave you to interpret it. It tells you what to do and records that it told you.
That shapes the software too. Most construction tools get built for a procurement committee instead of a foreman. Onboarding takes a week. The important button is three screens deep. The alert goes to someone who isn't on the roof. Auger has to work for a superintendent who never opens the settings screen. If it only works for people who configure it, it doesn't work.
Where things are
Auger is live on several active job sites in Florida, running crews through the back half of lightning season. Florida in August is the right place to find out whether this holds up. If it works there, it works.
Beta launch is October.
— John Loughrin