Mining
Space
UK Space Agency & Satellite Applications Catapult
SATELLITES FOR BATTERIES
PARTNERS
CLIENTS
REQUIREMENT
Can we combine the data from space with advanced analytics to help searching for the battery minerals?
CHALLENGE
Comparing highly uncertain historical data with the inputs from the disperse and different new data to produce accurate insights required finding a clever solution with the help of data science.
SOLUTION
A data processing and machine learning pipeline that produces prospectivity maps for multiple metals using a variety of inputs.
Bringing space and mining together
With transport being the UK’s biggest contributor to greenhouse gas emissions, the transition from conventional engines to Electric Vehicles (EVs) is vital but will bring an unprecedented increase in demand for battery metals and particularly lithium. Today mining companies are facing a huge challenge to meet growing demands, for EVs and the huge range of electronic goods that proliferate the modern world. The potential of lithium mining in Cornwall is high and this has been recognised by the government in its 10-year Integrated Defence and Security Review published in March 2021, which also highlights how important this is to the UK.
Is there a way that allows for a cost-effective and environmentally conscious exploration of the battery metal deposits? Surprisingly the answer lies not only on the ground, but also in space. The Satellite Applications Catapult is leading on a space and mining industry collaboration that uses satellite data to increase the identification of these metals both in the UK and internationally. Combining the data from space with advanced analytics will help to transfer revolutionary science and technology into world class capabilities, and help boost the UK economy.
The Pipeline to Prospectivity
The Satellite Applications Catapult invited Decision Lab to join its consortium made up of innovative mineral exploration company Cornish Lithium, leading space technology companies CGG Satellite Mapping, Terrabotics, and Pixalytics, and all supported by the University of Exeter’s Camborne School of Mines and the British Geological Survey. The solution from this collaboration was a data processing and machine learning pipeline that produces prospectivity maps for multiple metals using inputs from satellites and other remote sensing data. Gathering data in such a way and processing it through our AI based approach reduces the cost and environmental impact. The result is an AI model that helps to identify new target areas for battery metal extraction to meet the growing demand for critical minerals.
The Satellite Applications Catapult invited Decision Lab to join its consortium made up of innovative mineral exploration company Cornish Lithium, leading space technology companies CGG Satellite Mapping, Terrabotics, and Pixalytics, and all supported by the University of Exeter’s Camborne School of Mines and the British Geological Survey.
The solution from this collaboration was a data processing and machine learning pipeline that produces prospectivity maps for multiple metals using inputs from satellites and other remote sensing data. Gathering data in such a way and processing it through our AI based approach reduces the cost and environmental impact. The result is an AI model that helps to identify new target areas for battery metal extraction to meet the growing demand for critical minerals.
Training AI to learn the relationship between features?
Decision Lab’s focus was on developing the machine learning based method for producing battery metal prospectivity maps. We started by creating a “ground truth” for the existence of battery metals from historical mining data and mineral grab samples. This acted as a base line to which we would compare our findings about the location of the metals based on the results of our novel AI model. We could then start using the data provided to us by the project partners – lithology, geological faults, vegetation anomalies, thermal anomalies and land cover classifications – to train our machine learning algorithms. It learned to identify the relationship between the input data and the historic ground truth. We could then run the AI model against an area of Cornwall to produce prospectivity maps for multiple battery metals, including: Lithium, Copper, Cobalt and Nickel.
Mining for value
We had to overcome multiple challenges. There were many different types of geospatial data, at different resolutions and covering different areas, which we had to make consistent. Our historic ground truth was very limited, incomplete and highly uncertain – not the ideal training set. We had to be smart as to get the maximum accuracy with the limited training data and use methods that could cancel out the errors. The results have been very encouraging. The first test sample was collected from a location at which our model predicted a high probability of finding Lithium, but where experts would not have expected to find it. Analysis of surface rock found Lithium micas, which is a good indicator for the presence of Lithium.
Prospects for Satellite4Batteries
As Alastair Lees, Head of Extractive Industries at the Satellite Applications Catapult says: “This innovative and exciting collaboration between the space and mining industries will harness the insights provided by satellite data and advanced analytics to help identify potential locations of battery metals in Cornwall. By creating opportunities in the UK and internationally to improve the way mining companies locate these critical minerals, it will also support the UK’s green recovery and 2030 commitments to CO2 reduction.” We are looking forward to developing the technology further and seeing what the collaboration can achieve in this space, and the benefits it can provide in what is an internationally important endeavour.
Let’s collaborate
Send us an email, to discuss a new project.
We’re a team of innovators who are excited about unique ideas and help companies to create amazing solutions.