Satellite Imagery

The world of satellite imagery can be complex, confusing and difficult to navigate through. The traditional data provision model serves only a handful of expert customers. As well as bringing businesses easy access to satellite imagery, we’ve explained some of the basics hereafter, so that you can better understand this amazing source of insights.

the background

The first time the majority of people were exposed to satellite imagery was in the early 2000’s when Google launched GoogleEarth, the first public representation of Earth based on satellite imagery. Since then, there has been a huge increase in the number of mapping applications offering similar satellite basemaps, opening up new opportunities for the consumption and analysis of satellite imagery.

More recently, technology trends have opened up the access to space and enabled companies like DigitalGlobe, Airbus or Planet to launch Earth Observation satellites in greater numbers. By compiling and integrating this disparate data into our platform, we strive to offer one rich feed of the most valuable data at the economic scale (resolution under 1.5m) and temporal scale (at least weekly revisit), rather than a single, static and obsolete basemap.

On top of this, the latest technology advances in cloud infrastructure and artificial intelligence are being leveraged to extract unique types of insights at an unprecedented scale from this influx of data coming from space. We are calling this the democratisation of Earth Observation data.


some important metrics

 

Although the physics behind satellite imagery could be perceived as “rocket science” (i.e. something very difficult to understand); from a user perspective, it could be simplified and summarised in two words: it’s all about Resolution and Revisit.

The Resolution is the level of details you can see on an image. For example, in an image with a one-meter resolution, each pixel represents 1 meter on the ground.

The Revisit is the time you have to wait between two images being acquired for the same location. For instance, a daily revisit lets you monitor fast developing phenomenon.

From a satellite operator perspective, it means that either they have expensive and large satellites, usually in small numbers, with a large and sophisticated telescope acquiring very high resolution images (up to 30cm resolution for DigitalGlobe for instance); or they have smaller and cheaper satellites, hence in great numbers, with a small and simple telescope acquiring medium resolution images (between 3 to 5m resolution for Planet for instance).

Because the revisit is primarily driven by the number of satellites orbiting our planet, it means that operators of few but large satellites have a low revisit (between monthly to quarterly) while operators of large constellations of small satellites have a high revisit (between weekly to daily).

By combining images from a range of commercial satellite operators such as DigitalGlobe, Airbus and Planet, Bird.i offers the greatest spatio-temporal dataset, without compromising on the resolution or the revisit.

For example, the images above show the London Olympic Park at 0.5m and 5m resolution.


free vs. commercial satellite imagery

 

Accessing commercial satellite imagery remains a very difficult task with high commercial, contractual and technical barriers. Although it represents an invaluable source of insights for everyone, 9 out of 10 businesses use free online mapping applications such as Google Maps to make important decisions. Not only these businesses do not know how obsolete the satellite images are, but they are not aware that the current commercial Earth Observation satellite operators have the joint capacity to map every part of the world’s landmass more than once a day.

Free online mapping applications are based on advertising revenues and their target audience is where the world’s population is concentrated. As a result, their free satellite imagery is updated on average every 6 months – 7 years, depending upon the area. Over the top 100 urban centres of the world, this is closer to 6 months, over all other urban centres this is closer to 2 years; outside of these areas it is closer to 7 years, depending on the location’s remoteness and climate conditions. The map above summarises the average freshness of images from free providers.

For businesses that require access to the best and latest satellite imagery, a free service is unlikely to be the best source. However, getting hold of satellite imagery has been historically tricky and expensive due to a fragmented supply chain.
At Bird.i, we have solved this challenge by curating the best images from multiple commercial satellite providers and providing users access to our Image Service via a user friendly Portal or plug and play API.

 


earth observation and artificial intelligence

 

When looking at a location on a free online mapping application, or a satellite image if you know how to get one, you will see an area of interest at the time the image has been acquired, which might even be unknown. Depending on its resolution, it will show you what this location, and its amenities, looked like few months or few years ago. It would certainly then trigger some questions such as:

 

How was it before that image? What does it look like today? How will it look tomorrow?

 

These questions are very common and have been inspiring our Artificial Intelligence teams to develop smart processes which can learn from, and make predictions based upon, earth observation data. The aim of our AI work is to perform valuable analysis at scale, more precisely, and more quickly than with the trained human eye. To learn more about our intelligence services, click here.