Satellite imagery

Identifying Illegal Construction with UP42 and Simularity

Challenge

  • Easily accessing and analyzing high-quality satellite imagery of a property in Melbourne, Australia and its surrounding neighborhood during different time periods

  • Connecting Simularity’s Automatic Image Anomaly Detection System (AIADS) algorithm to large volumes of satellite imagery over time — in order to detect recent changes and possible illegal construction

Solution

  • Fast and simple access to high-resolution satellite imagery from Airbus’ Pleiades satellite using UP42

  • Analyzing the imagery to detect unusual changes occurring over time

Results

  • UP42 enabled Simularity to combine their powerful anomaly detection algorithm with high-resolution satellite imagery

  • Simularity were able to rapidly detect the demolition of the illegal construction on the property and additional anomalies that were previously unknown, in much shorter time than the time it would take someone viewing basic change detection tool output

  • Simularity made additional discoveries from the same imagery that were previously unknown, such as a new roof built recently on a nearby property

Background

Simularity Inc. is a San Francisco-based software company creating artificial intelligence technologies for satellite imagery analysis. Simularity’s mission is to create a future where we are able to react to the most important, sudden, and urgent changes and even predict them before they happen.

Simularity’s product, Automated Image Anomaly Detection System (AIADS) is an anomaly detection algorithm that enables changes to be detected over time. The algorithm sifts through vasts amount of geospatial data and flags hotspots of concern — anomaly hotspots that help customers find their needles in a haystack. Their customers are able to monitor their properties, rapidly detect changes, and benefit by generating revenue, saving costs, and catching compliancy issues before problems get worse.

However, it is difficult to estimate the actual number of illegal construction in Europe. For example, the Italian National Institute of Statistics warned of a “decisive rise in the level of illegality” in construction — involving nearly twenty new buildings in every hundred. In Greece, over 970,000 buildings were declared as unauthorized buildings since 2010. In Montenegro, over 80% of the houses and apartments fall under the term “illegal”.

In this particular case, one Melbourne, Australia property owner was in for a surprise when he found that the majority of structures on the land he lived on were illegally constructed by a previous owner over 20 years ago. Because prospective owners are expected to perform their own due diligence before purchasing properties, it became the owner’s responsibility to either gain the correct permits for the structures or to knock them down. After being quoted around $8000 to gain permits or demolish the illegal constructions — the owner faced an expensive dilemma.

Simularity aimed to look into this particular case, using AIADS in order to detect any recent changes to both the property and surrounding area.

Challenge

Easily accessing and analyzing high-quality satellite imagery of the property in Melbourne was difficult. With no easy way to combine Simularity’s powerful analytics with large volumes of imagery in a way that quickly demonstrated the power of their solution — Simularity were not able to use AIADS to detect changes in the property or neighborhood over time.

This meant that while Simularity had built a powerful analytics algorithm, they were not able to make anomaly discoveries in order to shed light onto the Melbourne property dilemma or potential future residential areas.

Solution

We provided Simularity with access to the UP42 platform. As a result, they had fast and simple access to high-resolution satellite imagery.

They selected and ran a data block using Airbus’ Pleaides satellite imagery — focusing on the Melbourne property’s neighborhood as the area of interest. Simularity’s aim was to detect recently changes to the property. This is why they set the date of analysis for AIADS to when any major changes occurred.

Results

UP42 enabled Simularity to combine AIADS — their powerful anomaly detection algorithm — with high-resolution satellite imagery. As a result, Simularity was able to detect the demolition of illegal construction, which was one of the home owner’s options.

Simularity also made additional discoveries that were previously unknown. Three major anomalies were detected from the same imagery — two trees were removed from a property, an additional roof was added to a house and a roof was changed on a house.

AIADS can now detect changes in residential properties such as anomalies caused by construction or demolition and the removal of trees. The algorithm is now used in order to provide a fast way to detect changes so that inspection can be carried out — within minutes of the satellite image being released. As a result, estimating and acting on the number of illegal constructions is now easier and quicker than ever before.

Using the AIADS algorithm on the UP42 platform opens new possibilities for city and regional municipal governments and related agencies and businesses to more rapidly detect important changes to their surrounding environments.

We continue to be an essential part of Simularity’s mission to detect rapid changes on the planet. Going forward, our collaboration is paving the way for a future where you will be able to pinpoint an area on Earth and receive an alert when something changes - no matter where you are.

Read more about this over on the Simularity blog.

Ready to get started?

Create your first workflow with 10,000 free credits, no credit card required. Questions? Check out our frequently asked questions.

Self-service access to geospatial data, flexible processing and analytics tools in the cloud