Poster Abstract: Cyber-Physical System for Collecting Data on Moisture Inside the Walls of Buildings Grzegorz Kłosowski Lublin University of Technology Lublin, Poland g.klosowski@pollub.pl Tomasz Rymarczyk. University of Economics and Innovation in Lublin Lublin, Poland tomasz@rymarczyk.com Marcin Kowalski University of Economics and Innovation in Lublin Lublin, Poland marcin.kowalski@wsei.lublin.pl ABSTRACT This paper presents the results of research on the identification of moisture inside the walls of buildings with the use of non-invasive electrical impedance tomography (EIT). The novelty and contribution of this research is the development of an original algorithmic method to solve the ill posedness, inverse problem. Since the new algorithm optimizes the method for each pixel of the tomographic image, taking into account a specific measurement vector, regardless of what and how many homogeneous methods are included in the algorithm, the obtained results are more accurate than those obtained with the use of homogeneous methods. As part of the research, prototypes of the EIT tomograph and electrodes for examining walls were designed and manufactured. CCS CONCEPTS • Computer systems organization → Embedded and cyber-physical systems • Network types → Cyber-physical networks KEYWORDS Electrical tomography, Machine learning, Moisture inspection, Dampness analysis, Nondestructive tests ACM Reference Format: Grzegorz Kłosowski, Tomasz Rymarczyk, and Marcin Kowalski. 2021. Poster Abstract: Cyber-Physical System for Collecting Data on Moisture Inside the Walls of Buildings. In The 19th ACM Conference on Embedded Networked Sensor Systems (SenSys ’21), November 15–17, 2021, Coimbra, Portugal. ACM, New York, NY, USA, 2 pages. https://doi.org/10.1145/3485730.3492868 1 Introduction Dampness in buildings is a serious threat to the health of people staying indoors. A humid environment promotes the growth of dangerous fungi, bacteria, and microorganisms. Water contained in the pores of the walls degrades them physically and chemically, and reduces the strength of building structures. All this leads to an increase in building maintenance costs. The aim of the research was to develop an effective, non-invasive system for detecting and imaging moisture inside walls [1]. Tomography is currently the only known technique that enables imaging of moisture inside the tested objects [2]. Due to the low resolution of the images and the difficulties in obtaining noise-free data, it is rarely used [3]. The research focused both on designing appropriate hardware and optimizing the method of transforming electrical measurements into tomographic images. Figure 1 shows the concept of an electrical impedance tomography (EIT) cyber-physical system for imaging the distribution of moisture inside the walls of buildings. Figure 1: The concept of a tomographic cyber-physical system for imaging the distribution of moisture inside the walls of buildings The possibility of obtaining a legible image of the distribution of moisture inside the walls is of great importance in preventing their undesirable effects. This study proposes a complete cyber-physical system, which includes an EIT tomograph with electrodes and software. Both the tomograph and the electrodes have been specially designed to measure the humidity of walls. An original hybrid machine learning algorithm was also developed that optimizes the output image using several homogeneous methods. 2 Materials and Methods The pixel-oriented hybrid method was used to transform the input measurements into pixels of the output tomographic image. The task of the above-mentioned method is to obtain a quality reconstruction that exceeds homogeneous methods such as, for example, elastic net, linear regression or artificial neural networks. All algorithms create pixel-by- pixel reconstructions, which means that a separate prediction model is trained for each pixel of an image. Then, using the classification algorithm, the best reconstruction method was selected for each pixel of the image, also taking into account a specific measurement case (input vector). Thanks to this, each measurement results in a different assignment of reconstructive methods to pixels. Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for third-party components of this work must be honored. For all other uses, contact the owner/author(s). SenSys '21, November 15–17, 2021, Coimbra, Portugal © 2021 Copyright held by the owner/author(s). ACM ISBN 978-1-4503-9097-2/21/11…$15.00 https://doi.org/10.1145/3485730.3492868 357