State of play

• Artificial intelligence is increasingly being brought to the source of data, thus establishing the Edge AI concept. Edge AI is gaining momentum across various industries and services by public authorities and enables new applications requiring high performance, ultra-low latency with high bandwidth, efficient power use, and intelligence beyond regular computing. Its strategic importance is emphasised through the EU Chips Act monetary investments in the EU semiconductor supply chain, highlighting the urgent need for a qualified workforce.

• The advancements in Edge AI applications face critical research and engineering challenges related to limited computing and energy resources of the edge devices, limitations of the existing AI algorithms and their insufficient fit.

• Also, the related rapid technological advancements, market dynamics, interdisciplinarity, and regulatory uncertainty set a unique challenge for the valorisation and management of Edge-AI innovations.

The general research objective of TIRAMISU is an effective and practical methodology and a corresponding framework for reliable and energy-efficient Edge AI hardware backbone design and innovation management. In particular, TIRAMISU sets the following four ambitious research objectives for the team of doctoral candidates:

❶ To develop the design approaches for three promising flavours of reliable and energy-efficient Edge-AI hardware accelerators for tomorrow.

❷ To introduce a methodology for Edge AI algorithm-hardware co-design and the automated deployment of Edge-AI algorithms onto relevant hardware platforms.

❸ To formulate and implement approaches for designing Edge-AI-efficient artificial neural network models with a focus on emerging ones supported by fitted encoding methods and bioinspired optimisation frameworks.

❹ To develop an approach for analysis of the role of human oversight in the organisation, exploration, and innovation management of the Edge-AI-based solutions design and to study system design for early-stage innovative AI-based technology development processes using individual research projects of TIRAMISU doctoral candidates in topics/WPs 1-3 as case studies.