August 15, 2022
The algorithm is based on the analysis of SAR radar images and detects changes in the frequency of the repeating wave that indicate the presence of lithium. During the proof of feasibility, it located three deposits in South America. “The model surprised us with its level of accuracy”
company Aster from the top of the eye (formerly Utilis), which specializes in detecting leaks from water infrastructures through the analysis of satellite images produced using the Synthetic-aperture radar – SAR, revealed today (Monday) a new algorithm based on depth learning, which allows to locate Underground lithium deposits through analysis of radar images from space. Lithium is one of the most sought-after raw materials today, mainly for batteries for electric vehicles, mobile devices and the chip industry. The price of lithium Amir 13 times in the last two years, And in March of this year, its price reached an all-time high of $77,000 per ton, and as a result lithium was nicknamed “white gold”.
the company Registered a patent on the new algorithm and is currently formulating a business model around the solution. In recent months, it has carried out a feasibility study in collaboration with two lithium exploration companies operating in Latin America. As part of the pilot, Astra located, on the basis of photographs received from radar satellites, three landmarks where there is a high probability of significant lithium deposits. Indeed, in these three sites lithium deposits were found at a commercial level. This success is noteworthy in light of the statistics in the mining industry, according to which the average lithium discovery rate is one in a thousand attempts.
find minerals from space
Unlike photography satellites, which are equipped with optical sensors operating in the visible light range and in the infrared spectrum, SAR satellites carry a radar operating in microwaves. Microwaves are able to penetrate through the clouds and deep into the earth. From the analysis of the characteristics of the wave returning to the radar, it is possible to learn about the properties of the soil, such as moisture and the composition of materials. Radar satellites orbit the earth in low orbit (LEO) and are used mainly for geophysical research.
The Astra company, established in 2015, uses the information captured by the radar satellites to locate minerals in the ground. Minerals, due to their electrical conductivity, cause changes in the frequency of the radar waves, and based on these changes it is possible to deduce which materials are in the soil, and in what quantity. To date, Astra has used this technology to detect leaks from water infrastructure. Locating the leaks is based on the difference between the composition of the substances found in drinking water compared to rainwater. When the system detects soil that is soaked with drinking water, the assumption is that it is a leak. In this area, the company works with more than 400 water corporations in more than 70 countries.
Lauren Guy, the company’s founder and VP of Technology, explained in a conversation with Techtime that there is a big challenge in analyzing SAR signals over populated areas. “The use of SAR satellites is widespread mainly in the security market, and less so in the civilian market, and this is because urban areas produce a ‘cloud of electromagnetic noise’, which is very difficult to filter. In fact, this is the core of our technology, the ability to use algorithms to filter out the noise.”
About a year ago, Astra established an independent team within the company, led by Guy, whose purpose was to examine additional uses of the technology. The focus on finding lithium is quite understandable, given the peak demand for the metal.
Guy: “There is a great demand for lithium, but the success rates in discovering lithium are very low, which makes the discovery of the mineral very expensive and also creates environmental damage. We started studying How lithium ‘looks’ on SAR satellites. Learning the machine was based on data analysis of lithium discovery companies that we work with. The algorithm we developed is also based on a combination of geophysical logics. The model demonstrated consistency and a very high rate of certainty, which even surprised us. The more information we collect, the more accurate it will be. In principle, we are able to scan the entire planet and locate lithium deposits.”
Astra, located in Rosh Ha’Ain, employs about 70 people. The company raised $6 million in March 2021 led by the Beringea Foundation and also has offices in the United States, Japan and the United Kingdom.
Posted in categories: big data, artificial intelligence, news
Posted in tags: Astra, Guy Lauren, SAR satellites, lithium
