A Greener Edge: A Framework on Carbon-aware Edge ML System Design

Edge devices are often deployed at scale, yet their environ- mental impact, shaped by complex interactions between hardware choices, workload demands, power systems, and deployment context, has been overlooked by the mobile com- puting community. We present MicroGreen, a design-time framework that models lifetime carbon emissions for edge ML systems. By combining component-level carbon mod- els with workload profiling and environment-aware energy analysis, MicroGreen identifies carbon-optimal configura- tions across diverse conditions. Our results show that the most energy-efficient processor is not always the most sus- tainable, and that ambient energy availability, inference rate, and deployment lifetime can shift the carbon-optimal design by over an order of magnitude. Through a real vision-based visitor detection and counting deployment in New York City parks, we demonstrate that heterogeneous, location-aware designs reduce total emissions by 47.26% compared to a ho- mogeneous baseline.

1The 24th ACM International Conference on Mobile Systems, Applications, and Services