Astro-Ecology group goal: By applying methods and technology from astrophysics, engineering and computer science, we aim to help solve major global challenges such as mass extinction, environmental pollution, climate change and disaster response (search and rescue).
An urgent global challenge: The world is currently facing the huge challenges of climate change, plastic pollution, industrial and domestic waste pollution, and the unintended impacts of natural resource exploitation by humans on the environment. As a result 60% of wildlife has disappeared over the past 30 years (WWF report 2018); putting us in the 6th mass extinction event in the history of life on Earth (the 5th being the end of the dinosaurs). We're also in an era where we have unprecedented potential to understand and address these challenges - and the need to do so has never been more urgent.
Astro-ecology approach: By understanding and effectively applying technology and physics-based approaches, we are assisting conservation and environmental action. Effective conservation relies on knowing the current state of the ecosystem being conserved. Response to environmental disaster requires knowledge of what contaminants are out there, and in what abundance and locations. By using drones we can get a bird's-eye view of the situation, and cover a large area on the ground very quickly. Using the best remote sensors for the task we are improving the efficiency and efficacy of human efforts to solve these crises.
Current research projects
Monitoring endangered animals and catching poachers: Thermal cameras allow us to see animals as a result of their body-heat. This negates their natural camouflage and allows us to detect them, and also poachers, in the dark. In thermal infrared footage animals appear as bright sources in a similar way to how stars and galaxies appear in astronomical imaging. We are applying techniques from astronomy and machine learning to detect and classify different animals automatically. We are developing a fully-automated drone system that can detect, identify and determine the health of animals. The system will be very low cost, robust, simple to operate and so user friendly that local communities in developing countries with no technical background can operate it independently. We are currently developing the system and starting to take this to different locations around the globe and, working with local rangers/researchers, helping monitor and project some of the most critically endangered animal species on the planet.
Combating climate change by putting out fires: Underground peat fires are one of the largest sources of carbon emissions globally. Peat is extremely flammable when dry, and in South East Asian nations peat fires can burn underground undetected for days, weeks, or months. The amount of carbon released by this burning can exceed the annual emissions from the global transport sector; and the fumes released cause millions of human health problems each year. Since they are underground, the fires are hard to detect. We have shown our thermal drones can detect the fires even though the ground. We are currently working with local fire fighting and conservation teams to build a system capable of automatically detecting these fires over large areas, so that they can be put out.
Locating and tracking environmental pollution: Finding and understanding chemical pollution usually involves collecting samples and analysing them in the lab later on. Using a drone-mounted spectrograph we are making this process more efficient, and mapping pollution over large areas. The spectrograph can pinpoint locations and allow us to identify chemicals remotely. By combining this with satellite data we can get the most complete picture of polluted areas to help environmental agencies deal with the contamination. We are testing this at disused mining sites around the UK.
Search and rescue: Similarly to detecting poachers, we are using our thermal drone system to find people in need of rescue. Drones and machine learning detection never get tired, and can be built to be very reliable. Using a drone allows search and rescue teams to find survivors quickly, whilst keeping the rescuers out of harm's way at the same time. We are building a system to save lives in partnership with Morecambe Bay Search and rescue.
Machine Learning working movies
|Prof. Steve Longmore||S.N.Longmore@ljmu.ac.uk||Finding innovative applications for astrophysics techniques outside of astronomy|
|Dr. Claire Burke||C.Burke@ljmu.ac.uk||Mission planning, physics of the environment, optimizing thermal data capture|
|Dr. Ross (Paul) McWhirter||P.R.McWhirter@ljmu.ac.uk||Machine learning, automated identification of animal species|
|Dr. Josh Veitch-Michaelis||J.L.VeitchMichaelis@ljmu.ac.uk||Computing and drone hardware integration|
|Dr. Marco Lam||C.Y.Lam@ljmu.ac.uk||Databasing, data flow and storage.|
|Maisie Rashman||M.F.Rashman@2016.ljmu.ac.uk||Quantifying and optimizing thermal camera performance|
|Non-ARI team members|
|Prof. Serge Wich||S.A.Wich@ljmu.ac.uk||Conservation biologist specializing in primates|
|Dr. Mara Mulero-Pazmany||M.C.MuleroPazmany@ljmu.ac.uk||Drone-conservation pioneer|
|Dr. Owen McAree||O.D.McAree@ljmu.ac.uk||Aeronautical engineer specializing in drones|
|Dr. Paul Fergus||P.Fergus@ljmu.ac.uk||Computer scientist and machine learning expert|
|Dr. Carl Chalmers||C.Chalmers@ljmu.ac.uk||Computer scientist and machine learning expert|