Several years ago, as a new Rivian owner, I was excited to go out on off-road adventures. To road trip, I had to do back-of-the-napkin calculations: how far I could safely drive between fast chargers, and which off-road locations I could reasonably reach from the nearest one. That one fast charger acted like a lifeline, and I had to estimate how far I could drive out and back to it. Exploring unmapped routes I hadn't accounted for would cost me extra battery %. This took a lot of the adventure out of the adventure, but added a new one: navigation and pen-and-paper range estimation.
My background is in mathematics and physics, and I've worked as a data scientist for many years. I invented a new machine learning algorithm specifically to do what I had been doing crudely and inefficiently by hand. It wasn't until more recently that I was able to develop it into a fully fledged application that is stable and functional.
The core algorithm I invented is unchanged. It's the piece I'm most proud of. It was an inspired idea: very simple, and robust enough for this kind of application.
I have now solved my own problem. I use the app daily, and the numbers on this page show exactly how reliable the model can be with plenty of driving. I use it to plan trips and have peace of mind about how far I can go and still get back.






