Case Studies

Our client is a US real estate management firm that helps its clients look for investment properties, i.e. buying vacant land with the best value. The process of looking for land worth the investment all over the country is a long and tedious process, which is why our client approached us with the task of developing an AI-powered system for analyzing satellite images and locating vacant land as well as detect various features.

The application analyzes satellite images of landscapes, detects land lot type, detects building and other structures, vegetation, checks for road access as well as analyses nearby lots to gain even more information about the lot.

The app determines if the land plots present in the satellite image are vacant or not by determining if buildings or other structures are present on the lot. For occupied land plots the app determines its type, e.g. commercial or residential, by classifying the buildings present on the property.

Vacant land plots are not created equal from the investment point of view, which is why our app analyzes the surrounding land plots to look for buildings and their type, thereby determining if there's water and electricity available for any future projects.

There are a number of important features for every lot of land which can dramatically influence whether the land is worth the investment, one of them being access to a road. Our app can detect if the property has road access from the satellite image.

Another important aspect is vegetation and tree coverage, which can influence how the land can be used. We have developed our app in such a way that it can determine if there's vegetation on the property and how much of the property is taken up by trees.

The land investment process, which previously took a lot of time and effort, is now fully automated. The investment firm can now analyze hundreds of land plots in a matter of minutes, reducing the decision time tenfold.