SEASIDE - Sustainable and robust AIoT for edge-based Services

SEASIDE is a research initiative targeting the environmental impact of AI and IoT, which has grown rapidly and risks diverging from IPCC climate targets. Rather than chasing performance at any energy cost, SEASIDE develops efficient, trustworthy ML models and distributed computing architectures that match state-of-the-art accuracy while drastically reducing energy use.

Its three core research questions ask whether it is possible to: (1) run complex ML on resource-constrained devices with minimal data, memory, and energy; (2) build distributed, collaborative frameworks that generate collective intelligence from local learning; and (3) create a general-purpose platform for experimenting with efficient, distributed ML without relying on large, energy-intensive datacentres.