
Machine learning is not as complicated as you might think, producing amazing results with minimal skill and resources
Finding a car park in the city is like finding a needle in a haystack. But not anymore.
Software developer Adam Geitgey recently commanded a plane surveillance camera to send him an SMS when street parking became available near his house.
Geitgey downloaded a free, open source model, already trained to recognise vehicles. He then commanded it to identify stationary cars – legitimate parking places, taken by cars.
Once a car leaves the space, he gets an SMS alert. Brilliant.
Geitgey writes a regular blog in Medium that explores the progress of machine learning. Its main message is this: as time passes by, one can easily perform tasks once considered very complex, or even impossible.


To set up this nifty SMS alert yourself still requires a certain level of programming knowledge. However, such tasks are more like a puzzle: get the pieces right, and it works.
You can imagine how much a customer, such as a city municipality, would have paid for this kind of application just a year ago!
Image copyright: Adam Geitgey, Medium