Evaluating the use of machine learning algorithms in environmental sensing for energy saving - Research report

Aug 4, 2026

Upload by Noah

Coastal lagoons function as highly complex ecosystems in which multiple interacting factors can significantly affect environmental quality and resource services. The SMARTLAGOON project integrates AI-based technologies with an efficient IoT sensing infrastructure to support the Mar Menor lagoon. The paper proposes a method to forecast chlorophyll and turbidity at future time instants using smart buoy data. Experimental results demonstrate that machine learning models can predict these sensed variables accurately, enabling energy-saving sensor operation.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages6

Reading Time15 min

AccessFree to download

Abstract

Introduction

Coastal lagoons and the Mar Menor context

SMARTLAGOON project and energy-saving motivation

Methods

Dataset description

Methodology and feature preparation

Results

Prediction accuracy of machine learning algorithms

Conclusion

Final remarks and future work

Evaluating the use of machine learning algorithms in environmental sensing for energy saving Giovanni Delnevo, Gianni Tumedei, Vittorio Ghini, Catia Prandi {giovanni.delnevo2,gianni.tumedei2,vittorio.ghini, [catia.prandi2}@unibo.it](catia.prandi2}@unibo.it)[ ](catia.prandi2}@unibo.it)Department of Computer Science and Engineering University of Bologna Cesena, Italy ABSTRACT Coastal lagoons are complex ecosystems characterized by the interaction of several actors, that can have a significant impact on them. The SMARTLAGOON project has the primary aim of integrating novel artificial intelligence-based technologies with an efficient Internet of Things (IoT) sensing infrastructure in the Mar Menor coastal lagoon. This paper presents an approach to predict some variables (chlorophyll and turbidity) usually sensed by the smart bouy in future instants of time. Results show that machine learning algorithms can accurately predict them. CCS CONCEPTS • Computing methodologies → Machine learning; • Computer systems organization → Sensor networks; Embedded systems. KEYWORDS environmental sensing, machine learning algorithms, coastal lagoons, digital twin ACM Reference Format: Giovanni Delnevo, Gianni Tumedei, Vittorio Ghini, Catia Prandi. 2023. Evaluating the use of machine learning algorithms in environmental sensing for energy saving. In The 2nd Workshop on Networked Sensing Systems for a Sustainable Society (NET4us ’23), October 6, 2023, Madrid, Spain. ACM, New York, NY, USA, 6 pages. [https://doi.org/10.1145/3615991.3616402](https://doi.org/10.1145/3615991.3616402) 1 INTRODUCTION Coastal lagoons are highly productive ecosystems with a number of different uses and services. Making up 13% of This work is licensed under a Creative Commons AttributionNonCommercial International 4.0 License. NET4us’23, October 6, 2023, Madrid, Spain © 2023 Copyright held by the owner/author(s) . ACM ISBN 979-8-4007-0365-2/23/10 . [https://doi.org/10.1145/3615991.3616402](https://doi.org/10.1145/3615991.3616402) the world’s coastline and approximately 5.3% of Europe [19], they are exploited for fishing, aquaculture, saltworks, and recreational activities, and play a crucial role in retaining and purifying pollutants, which is essential for the ecology of coastal areas [13] . Being affected by hydrological, hydrodynamic, ecological, and socioeconomic processes [6], these highly complex systems are particularly susceptible to both climate-related and human-induced pressures [1] . The Mar Menor coastal lagoon, situated in southeastern Spain, is the largest saltwater lagoon in Europe and is located in a region characterized by aridity and water scarcity [7] . The basin that drains into the Mar Menor experiences extensive land irrigation through water diversion from the near Tagus River, as well as excessive use of aquifers during drought periods [20] . Additionally, the Mar Menor attracts a significant population around it, as it relies on fishing and agriculture, as well as tourism during the summer season, as key economic activities [4, 9] . The rapid economic, social, and urban changes over the past several decades, along with historical mining impacts, have had numerous adverse effects on the ecological status of the Mar Menor, raising the need for urgent protective measures [10] . In this context, the SMARTLAGOON project, funded ina H2020 call, was developed with the primary aim of integrating novel artificial intelligence-based technologies with an efficient Internet of Things (IoT) sensing infrastructure. These technologies would gather input data for innovative socio-environmental dynamic models, enabling the forecasting of both short and long-term changes in the lagoon’s conditions. This information would then provide added value for management decisions aimed at safeguarding the ecosystem and services offered by the Mar Menor lagoon [2] . Not only that but the acquired knowledge can be applied to the realizat

FAQs

What problem does the document address in environmental monitoring?

It targets prediction of chlorophyll and turbidity in coastal lagoons to support more efficient operation of sensing systems and reduce energy usage.

How is the data collected for the machine learning approach?

The dataset is collected using a smart bouy deployed within the SMARTLAGOON project, with variables sensed at different time intervals and aggregated into averaged values.

Which variables are predicted using machine learning algorithms?

The study predicts chlorophyll and turbidity for future time instants based on sensor measurements.

Aug 4, 2026

Upload by Noah