(V1.0 August 2024)
As data analytics and AI models gain higher capabilities and provide new functionalities, their resource requirements (storage, power, energy) increase, making them less energy efficient. Moreover, the widespread accessibility to the tools based on those models leads to their extensive usage and a subsequent rise in their environmental impact. Therefore, it is crucial that the increase of algorithmic power is accompanied by appropriate measures and solutions aimed at concurrently maintaining or even reducing their ecological footprint for sustainability. This involves integrating sustainability by design and incorporating practices such as energy-aware training, constrained learning, etc.
On the other hand, Data and AI serve as primary catalysts for digital transformation, playing integral roles in numerous cutting-edge applications and products. Consequently, they can also emerge as pivotal forces fostering the development of novel, highly efficient energy technologies.
Therefore, this theme is planned to work in the following two directions:
- Rethink and evolve existing methods and tools, as well as design new ones, that, while keeping or increasing their algorithmic power, implies a reduction of their resources needs as well (data, computing power, storage capacity, etc.), thus resulting in better energy efficiency and reduced environmental footprint.
- Leverage AI and data technologies to combat climate change through energy-efficient and sustainable solutions, applications and use cases. That includes exploring and implementing AI-driven solutions in areas such as smart grids, precision agriculture, and environmental monitoring to create applications that optimise resource use and reduce carbon emissions.