Machine Learning-based Energy Optimisation - Internet of Things in Smart City - Optimized Temperature Monitoring

Aug 3, 2026

Upload by Finn

Deployment of Internet of Things (IoT) temperature sensors across urban areas is critical for understanding and monitoring thermal conditions, yet direct sunlight can overheat sensors and distort temperature readings. A machine learning approach is proposed to enable dynamic ventilation with small fans, improving measurement accuracy while reducing energy use. The study addresses steady-state temperature prediction from a limited initial measurement window and analyzes how ventilation time impacts both data quality and energy consumption. DNN models for low-power IoT devices are compared using multivariate time-series data.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages6

Reading Time15 min

AccessFree to download

Abstract

Introduction

Problem 1: Steady-state temperature prediction

Problem 2: Optimal ventilation time and tradeoff

Practical implications

Proposed approach and evaluation

Machine Learning-based Energy Optimisation Internet of Things in Smart City Eric Samikwa, Jakob Schärer, Torsten Braun, Antonio Di Maio Institute of Computer Science, University of Bern Switzerland source: [https://doi.org/10.48350/185798 | downloaded:](https://doi.org/10.48350/185798 | downloaded:) 4.6.2024 ABSTRACT The deployment of Internet of Things (IoT) temperature sensors in urban areas is essential for the monitoring and understanding of the thermal environment. However, accurate temperature measurements can be compromised by factors such as direct sunlight, leading to overheating and inaccurate readings. We propose a Machine Learning-based approach that addresses this challenge by dynamically ventilating the sensor environment using small fans, enabling accurate and energy-efficient temperature measurements. This paper focuses on two interconnected problems: predicting steady-state temperature using a limited window of initial temperature measurements and investigating the impact of ventilation time. We employ various DNNs suitable for low-power IoT sensor devices to predict temperature using multivariate time series from different sensors and compare their accuracy. Furthermore, we highlight the tradeoff between prediction accuracy, which is correlated to the length of the observed input sequence, and energy consumption dependent on ventilation time. By adopting advanced prediction techniques, we can develop efficient IoT systems for accurate and energy-efficient environment monitoring in smart cities. CCS CONCEPTS • Information systems → Sensor networks; • Computing methodologies → Neural networks; Distributed artificial intelligence; • Hardware → Temperature monitoring. KEYWORDS Machine Learning, Internet of Things, Energy Optimisation, Smart Sensors, Temperature Monitoring. ACM Reference Format: Eric Samikwa, Jakob Schärer, Torsten Braun, Antonio Di Maio. 2023. Machine Learning-based Energy Optimisation in Smart City Internet of Things. In The Twenty-fourth International Symposium on Theory, Algorithmic Foundations, and Protocol Design for Mobile Networks and Mobile Computing (MobiHoc ’23), October 23–26, 2023, Washington, DC, USA. ACM, New York, NY, USA, 6 pages. [https://doi.org/10.1145/3565287.3616527](https://doi.org/10.1145/3565287.3616527) 1 INTRODUCTION Deploying Internet of Things (IoT) devices, particularly temperature sensors, in urban areas has revolutionized our ability to monitor and understand the dynamic thermal environment. These sensors This work is licensed under a Creative Commons Attribution International 4.0 License. MobiHoc ’23, October 23–26, 2023, Washington, DC, USA © 2023 Copyright held by the owner/author(s) . ACM ISBN 978-1-4503-9926-5/23/10 . [https://doi.org/10.1145/3565287.3616527](https://doi.org/10.1145/3565287.3616527) play a crucial role in applications such as urban planning, building management, and environmental monitoring for smart cities [1] . However, their accuracy can be compromised by local factors such as direct sunlight, which can cause the sensors to overheat and yield inaccurate temperature readings. To address this challenge, we introduce a Machine Learning (ML)-based approach for ventilating the sensor environment using small fans, enabling accurate and energy-efficient temperature measurements. This paper focuses on two interconnected problems that arise in the context of the IoT temperature sensors. Firstly, we aim to predict the final temperature at a steady state after an extended ventilation time, leveraging only a limited window of temperature and humidity measurements taken during the initial ventilation phase. This approach enables us to conserve energy by activating the fan for a shorter duration, thus reducing the overall power consumption of the sensor device. Secondly, we delve into the issue of determining the optimal ventilation time that maximizes the amount of data observed and subsequently enhances the ac

FAQs

Why are IoT temperature sensor readings in smart cities often inaccurate?

Direct sunlight can overheat sensors, causing distorted measurements and inaccurate temperature readings.

How does the proposed method improve both accuracy and energy efficiency?

It uses machine learning to drive dynamic ventilation via small fans, enabling accurate temperature measurements while keeping fan runtime energy-efficient.

What is the main tradeoff studied in the paper?

Prediction accuracy improves with longer observed input sequences, but energy consumption increases with ventilation time, creating a balance between data richness and power use.

Aug 3, 2026

Upload by Finn