Carbon-Aware Scheduling for Cyber-Physical Systems in the Edge-Cloud Continuum

Sep 4, 2026

Uploaded by Theodora

Cyber-Physical Systems (CPS) enhanced with AI require dynamic compute allocation, yet rising demand increases energy use and carbon emissions. The work presents a carbon-aware scheduler for the Edge-Cloud Continuum that optimizes workload placement by jointly considering energy consumption, temporal carbon-intensity variations, and resource availability. It contributes a spatiotemporal scheduling algorithm and a discrete-event simulator using realistic MIT SuperCloud workloads. Results show major emission reductions via renewable prioritization and time-shifting, with caution that high cluster utilization can make time-shifting backfire.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages6

Reading Time15 min

AccessFree to download

Introduction

Edge-Cloud Continuum and carbon-aware scheduling motivation

Paper organization

Related Work

Carbon-aware workload scheduling approaches

Temporal shifting and spatial placement limits

Carbon-Aware Scheduling for Cyber-Physical Systems in the Edge-Cloud Continuum Johan Kristiansson Department of Computer Science, Electrical and Space Engineering Lule University of Technology Lule, Sweden [johan.kristiansson@ltu.se](johan.kristiansson@ltu.se) Jerker Delsing Department of Computer Science, Electrical and Space Engineering Lule University of Technology Lule, Sweden [jerker.delsing@ltu.se](jerker.delsing@ltu.se) Thomas Ohlson Timoudas Department of Computer Science RISE Research Institutes of Sweden Stockholm, Sweden [thomas.ohlson.timoudas@ri.se](thomas.ohlson.timoudas@ri.se) Abstract—Cyber-Physical Systems (CPS) powered by Arti􀀂cial Intelligence (AI) have the potential to revolutionize industries by enabling advanced analytics and autonomous decision-making. To support resource-intensive applications, there is often a need to dynamically allocate additional compute resources. The Edge-Cloud Continuum enables allocation and deployment of workloads across platforms, including IoT devices, edge clusters, and cloud environments. However, the growing computational demands of these systems can unfortunately result in increased energy consumption and higher carbon emissions. This paper investigates the development of a carbon-aware scheduler for the Edge-Cloud Continuum, designed to optimize workload placement by balancing energy consumption, temporal variations in carbon intensity and resource availability. Key contributions of the paper include a spatiotemporal scheduling algorithm, a discrete-event simulator capable of replaying realistic workloads from the MIT SuperCloud dataset, and a comprehensive empirical evaluation. Findings from the paper show substantial reductions in carbon emissions by prioritizing renewable energy sources and timeshifting workloads to periods of lower carbon intensity. However, when clusters operate under high utilization, time-shifting can inadvertently result in signi􀀂cantly higher emissions. In such scenarios, simpler greedy algorithms can be more effective. I. INTRODUCTION Cyber-Physical Systems (CPS) play an important role in modern society by seamlessly integrating computations with physical processes [1] . This integration enables real-time monitoring, automation, and decision-making across a wide range of industries. As these applications grow in complexity and scale, the demand for ef􀀂cient and scalable computational platforms becomes increasingly important. The rapid advancement in AI further underscores the need for new computing platforms that can meet growing computational demands. The Edge-Cloud Continuum [2] is a new paradigm in distributed computing that aims to unify edge and cloud computing. By dynamically distributing workloads, it becomes possible to optimize allocation of computational resources across platforms. For example, AI-driven CPS applications can process time-sensitive tasks on local edge servers and run resource-intensive tasks on remote cloud systems. However, as the demand for computing continues to grow, so does energy consumption. Data centers are expected to account for a signi􀀂cant share of global energy use [3], thus contributing to increased greenhouse gas emissions and accelerating climate change. Developing technologies to reduce carbon emissions is therefore critical to reduce environmental impacts. Lowering carbon emissions often correlates with energy ef􀀂ciency and costs. For example, scheduling jobs in regions or at times when renewable energy sources, such as solar or wind power sources, are available can reduce both costs and environmental impact, as renewable energy is typically cheaper than fossil-fuel-based alternatives. Micro-grids CPScan optimize renewable energy usage by delaying task execution until suf􀀂cient renewable energy is available. Edge platforms can process data locally to maximize the use of renewable energy and of􀀃oad execution to cloud resources in regions with lower carbon intensity, thereb

FAQs

What is the main goal of the proposed scheduler for the Edge-Cloud Continuum?

Minimize carbon emissions by optimizing workload placement while balancing energy consumption, carbon-intensity changes over time, and available resources.

How does the paper reduce emissions in its approach?

By prioritizing renewable-energy periods and timeshifting workloads to intervals with lower carbon intensity, combined with cross-platform placement across edge and cloud.

When can time-shifting increase emissions instead of decreasing them?

When clusters operate under high utilization, delaying execution can lead to significantly higher emissions, making greedy alternatives potentially more effective.

Sep 4, 2026

Uploaded by Theodora