Global Surface Intelligence

Global Surface Intelligence (GSi) is a data services company that provides machine learning and predictive analytics on large and complex data sets (‘Big Data’). GSi specialises in analysing satellite information and remote-sensing Earth Observations (EO) to solve customers’ business problems in new and novel ways.

This technology allows GSi to help its customers determine commercially valuable information on forest growth and crop yield predictions. GSi has developed bespoke machine-learning software that learns to recognise what satellites are “seeing” when covering important global assets such as forestry and crops. By understanding what is on the ground, often across vast areas, GSi provides stakeholders with invaluable ongoing insight into the value of those forests or crops.

The challenge 

GSi’s service proposition is founded on the GSi-Platform that collates and analyses vast amounts of satellite image data. The first phase of the platform was based on an in-house desktop computer. This was extremely inefficient in terms of time-to-results—a better platform architecture was required that provided both efficient data management and faster data analytics. 

How we helped

EPCC analysed GSi’s business requirements and examined the GSi-Platform workflow and software. We then moved GSi’s code from its desktop and laptop computers to our supercomputers at EPCC. This required some reengineering of the code and refining of the mathematics to enable it to run better and faster on our HPC systems. 

We also profiled the code to identify areas that could be fine-tuned to further improve software performance. By analysing GSi’s workflow, we found those areas where we could use parallel computing methods to run jobs to different processors (or cores) on our supercomputer, allowing many analysis tasks to be computed simultaneously.

The benefits

The resultant software is estimated to be 100,000 times quicker than other similar software when run on conventional computing systems, demonstrating the exponential benefits of running GSi software on HPC machines. As a result of this work, a full set of satellite imagery for the entire Earth’s surface can now be analysed in a few hours rather than days.

Further, GSi now has a commercial relationship with EPCC and utilises our HPC-platform as a service to provide the basis of its customer facing large-scale data management and analytics services.

The close working relationship with EPCC and GSi typifies a relevant and vital collaboration between industry and academia that helps a local SME tackle truly global challenges.