Empowering Edge AI with Systolic Arrays

The primary aim of EDGE-SA is to investigate the potential suitability and applications of Systolic Arrays (SA) and Evolvable Hardware (EHW) as acceleration solutions for data processing challenges that demand extensive computation at the edge, especially crucial for resource-constrained devices. These technologies can serve as supporting mechanisms alongside existing technologies like Artificial Neural Networks (ANN). Additionally, indoor tracking of assets is considered as a real-world application context to enhance and validate the research since, at present, there is no reliable commercial solution available in the market that can deliver high accuracy.