Non-destructive Diagnostics and prediction of damages in conventional monolithic materials is still a challenge. New materials and hybrid materials, e.g., fiber-metal laminates, pose hidden damages that show no externally visible change of the material. Well established measuring techniques are ultra-sonic monitoring and computer tomography using x-rays. Both techniques suffer from their high instrumental effort and difficulties in diagnostically robustness. External monitoring of internal damages of such materials and structures with simple and low-cost external sensors, e.g., strain-gauge sensors, under run-time conditions is of high interest. But there is a significant gap between knowledge and understanding of damage models and the interpretation of sensor data. Machine Learning (ML) is a promising method to derive sensor-damage relation models based on training data.

In this work, a multi-domain simulation study is presented comparing and evaluating different ML algorithms and models, i.e., Decision Trees (C45,ID3,ICE), Artificial Neural Networks (single and multi-layer), and many more. A simple plate is used as a Device under Test (DUT), which is modelled using a simple physical mass-spring network model (MSN), finally simulated (computed) by a multi-body physics engine. The physical computation of the DUT under varying load situations in real-time is directly performed by the simulator combined with an agent-based simulation of signal processing in a Distributed Sensor Network (DSN). Artificial strain-gauge sensors placed on the top of the DUT surface are computed directly from the MSN and are used to predict hidden damages (holes, inhomogenities, and impurities) by applying ML to unreliable sensor data. Monte Carlo simulation is used to introduce noise and sensor failures. There are some surprising results concerning robustness and usability of different ML algorithms showing that Deep Learning is not always suitable!

The simulator (SEJAM) combining physical and computational simulation can be used as a generic tool for investigation of ML, sensing, sensor design, and distributed data processing. The entire simulation can be controlled by the user via a chat dialogue controlled by an avatar agent. This feature enables the usage of the simulator for educational purposes, too.

There are two major scientific questions addressed in this work: (1) The suitability and accuracy of mass-spring models, especially for modelling of laminated and hybrid materials; (2) The suitability of ML for damage detection using noisy and unreliable low-cost sensors

A passive acoustic monitoring (PAM) device named SAMARUC has been developed to acquire
underwater sounds following the specifications of the Monitoring Guidance for Underwater Noise in European Seas: Monitoring Guidance Specifications. Based on a Texas Instruments processor, an ultra-low power ADC was programmed to work at a sampling rate of 192 kHz and adhoc electronics were designed allowing the processor’s two microSD buses to be used, thereby increasing the storage capacity. Many other software and hardware enhancements were implemented, such as the new low latency file system, the construction of NITUFF anodized aluminium housing and ringed buoys. With the resulting application, data obtained by the SAMARUC at El Gorguel (Cartagena, Spain) in 2018 were compared to a theoretical underwater noise map created using AIS data. This was done following the Descriptor D11.2 by means of the ambient noise level indicators at two one-third-octave frequency bands (63 Hz and 125 Hz), mainly related to marine traffic and noise pollution. The conjunction between the acquisition of underwater acoustic data and the development of a numerical propagation model was found to be highly recommendable to estimate the ambient continuous noise level when validating the acquired data as well as when correcting the prediction provided by the model.

In this study, a biocompatible and non-toxic pH-responsive composite film was prepared for food packaging application. The films are composed from polyvinyl alcohol as the main polymeric matrix, nanoclay as a reinforcing component and red cabbage extract as a non-toxic indicator. The prepared films showed lower water uptake values when the amount of nanoclay was increase up to 25 %. It was observed that the films become brittle at high loading of nanoclay (40%). The prepared films exhibited color change in alkaline and acidic medium due to the presence of red cabbage extract which turned pinkish in acidic medium and greenish in alkaline environment. The prepared films were characterized by FTIR and visible spectroscopy. The maximum absorption in acidic medium was (λ_max = 527 nm), while a red-shift occurred in the alkaline medium (λ_max = 614 nm). Future work will focus on crosslinking of the prepared films to improve their mechanical properties.

The design and usage capacity of mobile handheld devices have advanced tremendous in recent times, from being used solely for audio calls to the recent incorporation of augmented reality in smartphones. These new smartphone applications are power intensive causing excessive heating in phone parts, primarily batteries and processors. In some cases, the temperatures of phones in use exceed the 45 C threshold temperature of discomfort. These undesirable high temperatures affect user experience and form the basis of ongoing studies to improve thermal management in handheld devices. This work analyzes the thermal profiles of three smartphone models A, B and C for common tasks such as music playing, voice calling, video streaming and 3D online gaming. Transient surface temperature distributions were obtained with infrared imaging and thermocouple sensors, while processor and battery temperatures were obtained from inbuilt sensors, for the phones operating with and without phone casings (test and control cases, respectively). Test results showed that casings generally inhibit the dissipation of the heat generated within the phone, leading to increased processor temperatures. Comparisons between thermal profiles for different phone casing materials showed that plastic casings caused the least temperature difference throughout the test duration, followed by silicone, carbon fibre and leather casings, in ascending order. The rises in processor temperatures during the test duration were compared for the different smartphones: the phone B had the least temperature rise, followed by phone A and then phone C. These gave indications of the power consumption by the phone processors when undertaking the test tasks.

The 3D printing process deals with the production of three-dimensional objects with established geometries. However, this manufacturing process has a crucial point established at the beginning of the object manufacture, where anomalies can occur and compromise the entire object produced. The piezoelectric diaphragm has been studied as an alternative to the conventional Acoustic Emission (AE) sensor concerning the monitoring of structures and processes. It has in its assembling a ceramic element with piezoelectric properties, which makes its response sensitive to temperature variations. The Pencil Lead Break (PLB) method is widely used due to its efficiency in the characterization of AE sensors. The present work aims to study the influence of temperature on the piezoelectric diaphragm response for the monitoring of the 3D printing process. PLB tests were performed on the glass surface of a 3D printer at three different temperatures, and the raw signal was collected at 5 MHz sample rate. The signal was investigated in the time and frequency domain. The results demonstrate that the frequency response of the sensor is directly influenced by the temperature variations. In addition, the signal amplitude variations occur differently along the entire spectrum, and frequency bands with small and large amplitude variations can be selected for a comparison study. Furthermore, two frequency bands were carefully selected, and the mean error was obtained regarding the reference temperatures of 25 ºC and 45 ºC. It can be inferred that the piezoelectric transducer has low sensitivity to temperature variation if a proper frequency band is selected, where an acceptable error of 19.3 % was obtained.

Smartphone mode classification is essential to many applications, such as daily life monitoring, healthcare, and indoor positioning. In the latter, it was shown that knowledge of the smartphone location on pedestrians can improve the positioning accuracy. Most of the research conducted in this field is focused on pedestrian motion in a horizontal plane.

In this research, we use supervised machine learning techniques to recognize and classify the smartphone mode (text, talk, pocket and swing) while accounting for the movement up and down stairs. We distinguish between the going up and the down motion, each with four different smartphone modes, making 8 states in total. This classification is based on the use of an optimal set of sensors that varies according to battery life and the energy consumption of each sensor.

The classifier was trained and tested on a data set constructed from multiple user measurements (total of 94 minutes) to achieve robustness. This provided an accuracy of more than 90% in cross validation method and 91.5% if the texting mode is excluded. When considering only stairs motion, regardless of the direction, the accuracy improves to 97%. These results may assist many algorithms, mainly in pedestrian dead reckoning, in improving a variety of challenges such as speed and step length estimation and cumulative error reduction.

In general, water quality mapping is done by interpolation of in-situ measurement samples. Often, these parameters change with time. Due to geographic variability and lack of budget in Nepal, such measurements are done less often. Remote sensors which collect spectral information continually can be very useful in regular monitoring of water quality parameters. Landsat OLI bands have been used to estimate water quality parameters. In this work, we model two water quality parameters: Chlorophyll-a (Chl-a) and Dissolved Oxygen (DO) using Sequential Minimal Optimization Regression (SMOreg) which implements Support Vector Machine (SVM) algorithm and Recursive Partitioning Tree (REPTree) regressions. A total of 19 measurements were taken from Phewa Lake, Nepal and various secondary bands were derived from using Landsat 8 Operational Land Imager (OLI) bands. These bands undergo feature selection and regression models were created based on selected bands and sample data. The results showed satisfactory modelling of water quality parameters using Landsat 8 OLI bands in Phewa Lake. Due to a limited number of data cross-validation was done with 10 folds. SVM showed a better result than REPTree regression. For future studies, the performance can be further evaluated in large lakes with larger sample numbers and other water quality parameters.

Deep Learning strategies recently emerged as powerful tools for the characterization of heterogeneous materials. In this work, we discuss an approach for the multi-scale characterization of the mechanical response of polysilicon MEMS (micro electro-mechanical systems), based on data assimilation from two-dimensional stochastically representative images of the polycrystalline structure of films that typically represent the building block of the MEMS movable structures.

A dataset of microstructures is collected and a neural network is trained, to provide the appropriate scattering in the values of the overall stiffness (in terms e.g. of Young’s modulus) of the grain aggregate. Since results are framed within a stochastic procedure, the aim of the learning stage is not to accurately reproduce the microstructure-informed response of the polysilicon film, but instead to provide a fast method to be next used at the device level for statistical, Monte Carlo-like analyses of the relevant performance indices.

Accuracy of the proposed approach is assessed for different ratio between the dimension of the polycrystalline aggregate and the representative size of a single grain (i.e. for different number of grains gathered in the polycrystal), to check if size effects are correctly captured.

For several years, innovative and individualized therapy approaches with minor side effects using electroceuticals have been under development. In the future, electroceuticals will gain importance due to their broad range of applications in therapeutic approaches, such as against, chronic pain, rheumatoid arthritis, obstructive sleep apnea and cardiovascular diseases, to name but a few. Conventional electronic implants, such as hypoglossal nerve stimulators and vagus nerve stimulators, contain complex circuits composed of a large number of active electronic components, sensors and a voluminous battery unit. However, there is a need for miniaturization without impairing functionality and reliability to expand the field of application of electronic implants. Notable examples of these can be found among implantable microstimulators and biosensors. In this paper, tiny electronic implants with applications in functional electrostimulation are considered that contain neither batteries nor sensors or active electronic components. These are not intended for autonomous operation and require an extracorporeal wearable device. The utilization of intrinsic nonlinear properties of ferroelectric materials in ceramic capacitors could allow the implementation of sensor functionalities in microstimulators and tiny wearable devices. The focus of this work is on the use of these sensor functionalities for the development of a novel energy control concept. Energy is supplied via inductive coupling for frequencies below 1 MHz. A mathematical model was implemented using Mathcad Prime 3.1. This nonlinear model includes the hysteresis modeling of ferroelectric materials. For model validation, comparative calculations were performed with ANSYS 2019 R2.

Electromagnetic (EM) characterization of engineered artificial materials such as bio- and nanomaterials is very important for several reasons. One of these reasons is future proliferation of 5G communication networks exposing urban population into the EM radiation of wide spectral range, therefore it is critically important to understand how new materials EM response can be utilized in electronic and communication devices and also ensure EM compatibility of biomaterials used inside human body. A new method to characterize permittivity and permeability of artificial materials using capacitively loaded aperture sensors is proposed and experimentally evaluated. The advantage of this new method over the existing techniques (free space, loaded waveguide, microstrip and coplanar waveguide resonators, coaxial probe, etc) is three-fold: i) resonance EM field enhancement inside the loaded aperture leads to very high sensitivity and therefore accuracy of EM parameters de-embedding; ii) only small thin samples of material-under test are required (with sample area substantially smaller than squared wavelength of radiation); iii) the method is easily scalable over the frequency and wavelength and based on relatively simple permittivity and permeability de-embedding procedure. This method can find application in EM materials characterization for existing and future electronic and communication devices.