Google DeepMind’s GraphCast AI Weather Forecasting Technology

Pictured above is Jacklynn Stott receiving the Google DeepMind Certificate at the Worshipful Company of Engineers and Engineers Trust Awards Dinner on 3 June 2025 from Dr Shini Somara with Master Engineer Eur Ing Penny Taylor, JP and the Clerk [photo courtesy of Mark Witter Photography] 

Google DeepMind: GraphCast, A step change in weather forecasting 

GraphCast represents a revolution in weather forecasting. Its AI-powered technology uses cutting edge machine-learning algorithms and vast data sets to provide highly accurate and more timely weather predictions that, among other advantages, could help mitigate the impact of severe weather events and, ultimately, save lives.

Instead of the traditional numerical weather prediction methodologies, GraphCast uses machine learning to provide highly accurate and timely weather predictions up to 10 days in advance. It harnesses the power of graph neural networks to model complex weather patterns and offers unprecedented precision in forecasts and an enhanced ability to predict extreme weather conditions. The model currently takes just 45 seconds to generate a forecast that would take traditional forecasting methods over an hour on a supercomputer. This has been equated to a five-year leap forward in weather forecasting.

By significantly improving the speed and reliability of weather predictions, GraphCast has the potential to support critical decision-making across various industries, optimise resource allocation and help mitigate impacts of severe weather events, allowing authorities to issue safety and evacuation warnings ahead of time and potentially save lives.

The Royal Academy of Engineering MacRobert Award is the premier prize for UK innovation in engineering. It seeks to demonstrate the importance of engineering and the role of engineers and scientists in contributing to national prosperity and international prestige. It is awarded annually for an outstanding example of innovation and benefit to the community, which has also achieved commercial success. Each year the winning team receives a gold medal, widespread publicity, a £50,000 prize and an exclusive weekend away at Douneside House. 

Originally founded by the MacRobert Trust, the Award is now presented and run by the Royal Academy of Engineering. The Engineers Trust is supporting the Award with £20,000 annually for 10 years. The 2024 Award was first announced at the RAEng Awards Dinner on 9 July 2024.

The Master Engineer & 2024 MacRobert Award Winners Google Deepmind
The Master Engineer Dr Dolores Byrne OBE and 2024 MacRobert Award Winners Google Deepmind at the RAEng Awards dinner on 9 July 2024 (photo credit Rob Lacey)

The MacRobert Award winning Google DeepMind team members are:

  • Ferran Alet, Research Scientist
  • Peter Battaglia, Research Scientist
  • Meire Fortunato, Staff Research Scientist
  • Remi Lam, Staff Research Scientist
  • Shakir Mohamed, Director of Research
  • Alexander Pritzel, Senior Staff Research Scientist
  • Alvaro Sanchez-Gonzalez, Staff Research Engineer
  • Jacklynn Stott, Research Program Manager
  • Matthew Willson, Staff Research Engineer
  • Peter Wirnsberger, Staff Research Scientist

The other 2024 MacRobert Award finalists were:

Sunamp: Transforming energy storage to combat climate change

The Master Engineer & 2024 MacRobert Award finalists Sunamp (photo credit Rob Lacey)
The Master Engineer Dr Dolores Byrne, OBE and 2024 MacRobert Award finalists Sunamp (photo credit Rob Lacey)

Sunamp has pioneered the development of innovative heat batteries utilising phase change materials to revolutionise thermal energy storage. These advanced batteries store and release large amounts of thermal energy through the melting and solidifying of the specially developed phase change materials, providing a highly efficient and sustainable solution for heating and cooling applications. Current products provide heat storage for domestic hot water that are more space efficient and energy efficient than traditional hot water cylinders.

University of Oxford, AstraZeneca and partnersScale up and manufacture of ChAdOx1 Covid vaccine

The Master Engineer & 2024 MacRobert Award finalists University of Oxford, Astra Zeneca & partners
The Master Engineer Dr Dolores Byrne OBE and 2024 MacRobert Award finalists University of Oxford, Astra Zeneca & partners (photo credit Rob Lacey)

The University of Oxford and AstraZeneca led a consortium of manufacturers, suppliers and other partners in the development of an innovative manufacturing process that was rolled out globally to supply over 3 billion doses of the ChAdOx1 Covid vaccine, saving over 6 million lives.

Royal Academy of Engineering Announcement

The MacRobert Trust website

The Worshipful Company of Engineers Charitable Trust (the Engineers Trust) acknowledges excellence in engineering, supports engineering education and research, gives grants and assists in the relief of poverty.