Opportunities
- According to the ICT Environmental impact 2024 report of the Rolling Plan for ICT standardization, ICT technologies and digitalisation can help reduce 15-20% of total greenhouse gas emissions.
- The green transition presents a major opportunity for European industry by creating markets for clean technologies and products, and the European Green Deal is the instrument to make it possible.
- The European climate law, with the legally binding target of net zero greenhouse gas emissions by 2050, translates previous objectives into law and emerges as an opportunity to address at EU-level the reduction of greenhouse gas emissions.
- The Green Deal Data space can set up the foundations for data sharing to address the challenges related to climate change and efficient use of resources.
- Investments in greener technology, e.g., EuroHPC green supercomputers.
- Focus on the broad deployment and effective implementation of an EU circular economy and DPPs.
- Invest in Quantum computing to increase energy efficiency by an order of magnitude
Priorities for BDVA
- Create awareness of the impact technologies such as GenAI have on the environment and foster a responsible use of such technologies.
- Contribute to the application of Data and AI for the design of environmentally sustainable solutions. Foster the development, adoption, and usage of tools and practices especially aimed at improving energy and resource efficiency.
-
- Data analytics tools to monitor and reduce the environmental impact of data centres and data processing, aligning with sustainability goals. Implement and advocate for sustainable data processing and storage practices, focusing on lowering data centres’ environmental impact.
- Sustainable process optimisations.
- SW and modelling tools that can assist designers and developers in considering factors like energy efficiency, waste reduction, and green spaces
- Contribute to developing specific energy-efficient AI techniques or approaches based on current and future lines of discussion at a technical level. This set of techniques can include, among others neuromorphic computing and spiking neural networks; energy efficiency approaches in IT (AI, training, computing) combined with vertical perspective (holistic view); surrogate models to reduce computational load, complexity and power consumption while maintaining an acceptable level of accuracy; foundation models specifically designed for vertical sectors, so less data/energy is required (reduced model sizes, modular and designed for specific domains); green AI (energy aware training and learning, specific metrics) and more compact GenAI models, combined with specialised processors, and the increasing adoption of quantum computing; constrained conditional AI models; frugal AI; a mixture of experts (MoE); AI quantisation; small data; edge learning.