Skip to main content

The art of more

By [email protected] - 17th May 2017 - 17:26

We have been creating imagery of the ground from above the ground since 1858, when Gaspard-Félix Tournachon loaded an entire darkroom into the basket of his Montgolfier and while tethered 80m above the French village of Petit-Bicêtre, took the world’s first aerial photograph.

Since Tournachon’s afternoon of hanging about, a huge amount of earth observation data has been created at an ever-increasing rate. The most recent development has seen much money being invested in new companies that are developing, building and planning to launch swarms of satellites, which will create even more EO data. Setting aside this new and slightly ‘Call of Duty’ collective noun, it can only be good news that this section of the geo-industry is attracting so much interest.

But those same people with huge piles of cash in the bank and who want to gamble on making more are now slowly turning their attention not just towards companies that create imagery, but those planning to do new, interesting and (crucially) valuable things with these images.

Some of these start-ups are well known – I would put Orbital Insight and Descartes Lab into this category – but many are, to shoehorn in an awful pun, flying below the radar. I will leave the obvious reference to the importance of marketing hanging there and move on to look at some companies I hadn’t heard about until I started looking.

London-based Terrabotics advertises that it can take the ‘pulse of the planet’ by deploying proprietary algorithms to convert remote sensing imagery into ‘actionable intelligence’. Terrabotics’s founders claim 10 years of research have gone into its ability to automatically create terrain models from the vast amount of satellite, aerial and drone imagery available. Being able to accurately model and measure an area of interest that may be difficult and hence expensive to access will certainly bring benefits to a wide range of industries.

Like Terrabotics, Spaceknow will help its customers to gain ‘real-life insights into economic activities’ of their areas of interest. At the ripe old age of four, this Prague and Silicon Valley co-located start-up recently raised US$4 million to support the continued development of its deep-learning algorithms and delivery platform. Two customers of that platform are Chinese and African financial indexes, which use it to learn more about economic activity in their respective regions and provide a check on the ‘official’ figures.

Bluefield is one of the ‘future-swarm-launching’ companies, but its choice of sensor and promise to provide ‘actionable datasets’ means that it qualifies for this list (hey, it’s my list!). From its offices north of Los Angeles, the company will draw down the data it needs from its gazillion 12kg CubeSats to monitor the methane outputs on Earth. I don’t know if its claim to be able to detect 0.6kg/hour means that cows across the globe will be for evermore embarrassed by technology, but it is revealing that one of their target markets is hedge funds (no pun intended). If actionable intelligence equates to gaining the market, then people are going to pay for it.

All these companies are academic in their development history, sensible in creating intelligence not imagery (especially when the price of imagery is under huge price pressures) and ambitious in targeting new but rich customers.

There will always be end-users of remotely sensed data that simply want the data delivered so that they are able to use their skills and experience to complete their work but the number of those users is unlikely to grow in a commercially meaningful way.

Mass market users of any technology don’t care what that technology can do, only what it can do for them. By building companies around that truism, these new providers of ‘earth intelligence’ could just burst out of their metaphorical garages and become the future titans of remote sensing.

Alistair Maclenan is founder of the geospatial B2B marketing agency Quarry One Eleven (

Download a PDF of this article