Satellite imagery has become an indispensable part of our lives, and has affected it in more than one ways. If you are wondering how—well, touring exciting places on Google earth, telling us how much the agriculture produce would be, how much profit a supermarket is going to make or most recently what could be the actual oil content in a country, is all thanks to satellite imagery and its analysis! Uncovering a well kept secret such as the oil stores in a country may seem quite astonishing, but that’s exactly what Silicon Valley startup Orbital Insight has done by applying machine learning algorithms to put the pieces of data together– which are nothing but pictures and raw data collected by the satellites orbiting the planet.
Satellite imagery is possible because of two parts in it—those who launch satellite and those who analyse the satellite data. The focus of analytics focused organizations in machine learning and AI applicability, is what results in the awesome pieces of information that we see. The catch here is not the fancy images these satellites can produce, but the large data sets that can yield insights into activities like supply chain, agriculture, oil production etc.
Commodity Price indexing with satellite imagery—from retail to agriculture
There are many startups booming up in this space who are scanning satellite photographs and combining machine learning knowledge to identify trends in various fields including agriculture, mining etc. By using satellite imagery, it has become possible to count the number of cars in the store which can determine factors such as customer flow, busiest times for retailers, which stores are being crowded the most by people and more. For instance, Orbital Insight can track more than 90 major retailers in the US. The analysis based on this information can help in deciding commodity pricing in the retail industry based on these observations.
For instance, a satellite intelligence by a firm called RS Metrics on JCPenney’s parking lots across US could tell that in the month of April and June, the traffic into the stores were on rise. This sent out alerts to its clients in the real-time, enabling the them to trade on these numbers.
Finding out the agricultural output is another interesting tracking satellite imagery analytics could do. It is done primarily by tracking the infrared light that these plants emit. It is interesting to note that the chlorophyll content of these plants emit infrared light during the process of conversion of sunlight to energy. It is based on this activity, that it can be determine if a plant is healthy and what could be its probable productivity. Infrared sensors combined with satellite images can prove to be a boon for commodity traders who want to bet high on the crop yield.
There are even few companies who have been using earth imagery to find out the investment opportunities. Organizations and traders have been relying on the real-time or near real-time satellite data to fetch elusive information edge, thanks to alternative data—which is nothing but raw or unstructured data that is distinct from company filings, historic market prices and similar entities.
Uncovering secret oil reserves in China with satellite imaging-
The US based Orbital Insight came up with precise data on rate of production and consumption of crude oil in China. After developing a new way to calculate China’s above-ground oil reserves, using machine learning and computer vision, they could not only detect and identify China’s oil tanks but also measured the overall supply.
Since crude oil prices are based on the net supply and consumption, it is important for investors to have the most up-to-date and accurate data as possible. The data by Orbital Insight found out ample reserves that China has. It revealed a possible fudgy numbers by the Chinese government to show inflated prices. This shows how oil producers made a way to have high margins.
Is India pacing up with satellite imagery?
As India is pacing up in its approach of satellite launches, it hints towards an increased interest and reliability on satellite images to get the best data in every field. During the launch of GSAT-9, Prime Minister Narendra Modi had mentioned about how the South Asia Satellite would go a long way in addressing the region’s economic and developmental priorities. He is hopeful of carrying out natural resource mapping, telemedicine, education, deeper IT connectivity with it.
Another startup, SatSure, a data analytics firm combining satellite, weather, and IoT analytics with the agriculture sector, is helping in governing crop growth and soil health. Using its proprietary machine learning and parallel computing techniques, it is aiming to resolve complex relationships and get insights into the crop phenology. Apart from ensuring an immediate compensation for crop loss due to adverse climatic conditions, satellite intelligence can help traders in setting the price yields based on its produce.
Though at a steady pace, India is catching up in utilizing satellite imagery in using across all verticals of trading, which is evident from the continuous efforts by the government to launch satellites and analytics based start-ups cropping in to make the most of the data generated from these.
Try deep learning using MATLAB