Data science and AI

The North East Space Communications Accelerator (NESCA) brings together academic research with business expertise to drive growth.

The accelerator builds on investment in space skills research and innovation in the North East of England. It aims to boost knowledge exchange and impact, create highly-skilled jobs, and drive economic growth.

The project is a collaboration between Northumbria, Durham, and Newcastle universities, the North East Combined Authority, Space North East England, and the North East Space Leadership Group. There are 14 industry partners invovled too. We are leading the AI in Space research theme and skills-related activities.

We're developing a cutting-edge crop evaluation toolkit.

We're integrating field monitoring, phenotyping, machine learning, and early stress detection. The combination enables more precise and proactive crop management.

Applications include real-time stress detection, advanced physio-genetic analysis to pinpoint infection hotspots, and the screening of stress tolerant genotypes.

By characterising key morphological traits, the toolkit

• helps tackle issues that can stress crop plants
• enhances crop resilience
• improves yields
• supports sustainable agricultural practices

Advances in AI, fuelled by big multi-sensor data, are revolutionising how systems operate and enabling intelligent decision-making. However, these advancements are often data-driven and lack full transparency in decision-making processes. 

We are leading the development of explainable AI-driven, multi-sensor data solutions to support autonomous systems and, enabling cutting-edge applications, including: 

• multimodal wide area surveillance: enhancing situational awareness across large geographic areas in earth observations, and space monitoring with satellite multimodal information.
• multi-task learning for maritime data: improving performance in maritime monitoring tasks using geo-located multimodal datasets.
• anomaly detection for multi-target analysis: advancing detection, localisation, tracking, and behavioural analysis to identify unusual patterns, activities, or threats effectively. 

These innovations have potential to improve security, safety, and operational efficiency in Earth observation, multi-sensor satellite data, maritime operations, and anomaly detection.

We’re delivering the world’s first MRes and PhD research into countering deepfake satellite imagery. Conducted within the Geospatial Systems Centre for Doctoral Training (CDT), this pioneering project aims to:

• investigate AI-driven image manipulation and build a reference dataset of fake images.
• develop AI-based techniques to detect manipulation and verify media integrity and authenticity.
• provide tools to counter AI-driven threats, ensuring a resilient and trustworthy geospatial intelligence ecosystem.

In an era of digital deception, this research strengthens global security by safeguarding satellite imagery integrity.

Earth observation satellites currently collect data and send it to ground stations. Then it is processed in data centres or the cloud.

This approach generates large volumes of data, leading to considerable operating cost inefficiencies, and delays in decision-making.

We are exploring new ways to improve Earth observation systems. By using modern satellite constellations, AI, and advanced communications, the goal is to enable satellites to process images on-board, analyse data, and share information directly from space, pioneering a smarter and more efficient EO ecosystem.