Professor Plamen Angelov, Chair in Intelligent Systems and founding Director of the Lancaster Intelligent, Robotic and Autonomous systems (LIRA) Centre, has been appointed Visiting Expert for the European Space Agency’s Phi School.

Professor Angelov, who has over three decades of experience in advanced research on knowledge and human-interpretable machine learning models, received the appointment as a result of several years’ of collaboration with the European Space Agency’s Φ-lab, in which he spearheaded a project that would use artificial intelligence (AI) systems to predict and warn of flooding. With the Φ-lab, he developed a new AI system that was able to efficiently analyse and interpret streams of satellite imagery and other data to improve flood alerts and procedures for rescue planning. His post as Visiting Expert will continue until August 2027.

The project, known as Explainable AI for Earth Observation (AI4EO), endeavoured to make transparent the workings of AI in this field, ensuring that any conclusions drawn by AI are based upon sound, robust data, and would be easily interpretable for human analysts to make decisions based upon its suggestions. A significant part of the AI4EO programme involved the Sentinel-2 mission, a European Space Agency satellite tasked with gathering detailed data of global land coverage from 56° S to 84° N, including coastal waters and all of the Mediterranean Sea, using multi-spectral data with 13 bands in the visible, near infrared, and short wave infrared part of the light spectrum.

Using data gathered from this satellite, Professor Angelov and his team utilised modelling software to assign each pixel within the satellite maps a class, flagging whether it was land, water, or cloud, and colour-coding it as to the confidence of each estimate. When the satellite data produced these prediction maps, the team would then be able to compare them with “historical” water records. Where there were significant differences between the historical records and the satellite-derived ones, this could be indicative of areas at risk of flooding, allowing flood prevention teams to investigate.

On his appointment to the Visiting Expert position, Professor Angelov said: “Joining the Collaborative Innovation Network of the European Space Agency’s Φ-lab and its Phi School whose motto is “Machine Intelligence meets human creativity” is a great opportunity to contribute to, collaborate and interact with top researchers whose goal it is to make best use of the rich and ever-growing data sets and streams resulting from Earth Observation under the Copernicus Programme and its Sentinel missions. It is a privilege to help inform decision makers and the public about the adverse climate events such as flooding, and to get the opportunity to make a tangible difference to communities in this way. Φ-lab’s and my personal interests are primarily around the research and use of explainable and continual deep learning form of AI, and I look forward to working with them closer in the coming years."