What we do

We have a growing portfolio of work applying forecasting and optimisation techniques to energy sector challenges. Project examples include the evaluation of forecast accuracy and weather regimes on the economic dispatch and spinning reserve requirements of fossil fuel generators, analysis of factors driving unscheduled downtime for a wind farm and the optimisation of production to minimise energy costs and CO2 emissions.

Our applied meteorologists work operationally with data from global forecast providers and run our own in-house WRF meso-scale weather model. We are experts in sourcing and manipulating observational data, forecast verification and performance benchmarking. We have practical experience working with climate reanalysis data and climate model outputs and have developed several applications for ECMWF’s Climate Change Service and Data Store.

We have many years of experience in data visualisation, with a particular focus on integrating geospatial data with other information sources, including scientific and socio-economic data. We combine a deep knowledge of the scientific principles behind the data with an appreciation of the need to present data in an engaging and meaningful manner that communicates a clear message or enables better decision-making.

We have a wealth of expertise in the development and deployment of data-driven models of virtually every conceivable type. Specific methodological expertise includes deep neural networks, support vector machines, generalised linear models and Bayesian belief networks. We also frequently employ optimisation techniques such as mixed integer programming to identify optimum decision thresholds.

We have many years of experience across all areas of software development, ranging from small stand-alone apps to full-scale commercial SaaS products. Our front-end team writes modern, single page applications which enable intuitive data exploration leveraging our in-house UX design skills. Our back-end team develops GIS-enabled services exploiting the latest containerisation technology and cloud services to deliver secure and resilient performance.

We have significant experience in processing many kinds of satellite imagery (optical, radar, thermal and others) using image analysis and machine learning algorithms. Our projects have included studies of land cover classification, urban change detection, atmospheric composition (e.g. air quality and pollutant analysis) and surface temperature using Earth observation data from both free and commercial sources.