Bigfootprint
Over the past few months, I’ve been watching a massive Amazon distribution warehouse being constructed in Thornton Colorado. These high resolution satellite images span from April — Dec 2017:
This structure is right off the freeway and in proximity to other “big box” stores like Target. Amazon’s box dwarfs Target’s.
This is a huge building, it’s footprint is about 80,000 m2, but it doesn’t even crack the top 20 buildings in the world by footprint size!!
Let’s see what these look like from space.
(random side note, a pet peeve of mine is when people say “it’s so big it’s visible from space” — DigitalGlobe satellites can see objects 30cm wide from space)
When buildings of this scale are built, it is very obvious when looking at satellite imagery. But what about smaller changes? Like a 10ft x 10ft storefront, or road widening? The latest algorithms and remote sensing technology allow end users to avoid processing data altogether and get right to change information. This breakthrough is made possible by massively storage, compute and machine learning with tools like SageMaker:
https://aws.amazon.com/solutions/case-studies/digitalglobe-machine-learning/
Now that DigitalGlobe is part of Maxar Technologies, we can detect these changes using Radar in addition to images. Radar is an active system which broadcasts waves and measures their return. This means it can be done at nighttime, or when clouds are obscuring the ground (which is the case 70% of the time).