ππ Today, I want to share a perfect use case for AI and how a naive solution without it failed me! π»
Things you need to know about me:
But hold up... why is navigating Montevideo on two wheels a tech-tangle? π§πΊοΈ
Neither #GoogleMaps nor #AppleMaps are up to speed with #BikeRoutes in my city... Talk about a modern-day muddle! π€ Using pedestrian or car options? Sure, if you don't mind wonky timings and a lack of elevation awareness π΅ββοΈ
My solution? I rigged up a DIY iOS shortcut to gauge my ETA to any spot in the city π, factoring in the good ol' crow's flight distance and my average pedal-pushing pace π΄ββοΈπ¨. But clearly, it's not cutting it.
So, do I need AI? You bet I do! The basic linear model: average speed βοΈ distance isn't good enough π οΈβ‘οΈπ§
Here's why: π I often take the path along the coastline, much longer than the great circle distance. ποΈ But, sometimes, I move "in-land," and the estimate is accurate. π¬οΈ Wind? It slows me down a lot. βοΈ Weather conditions? Major vibe-check for my speed. ποΈ Elevation changes? Add another few minutes to the ride.
I'm on the brink of training an #XGBoost regression model to nail down my travel times ποΈββοΈπ. Given the start/end GPS points and the current weather, I'm counting on more precise predictions! π―
The catch? It will only work for Montevideo; it will not be ready to take on the world... yet. πβ
I'm always down to hear how others are using AI in their day-to-day lives! Share away in the comments! ππ¨βπ»