Search

This paper proposes a specific domotic sensor network to measure the well-being of elderly people in private home environments through Machine Learning (ML) algorithms trained with daily surveys. The tests have been conducted in 5 apartments lived by 8 older people where the non-obtrusive sensor network is installed. Two ML algorithms are compared, Random Forest (RF) and Regression Tree (RT), such that to verify whether the users’ well-being is encoded in behavioural patterns obtained from the domotic data. These data are used to measure users’ well-being and compared with three reference indices obtained through a daily survey: a physical (Phy), a mental (Mind) and a general health index (Avg). The extracted indices from the daily survey are used to train ML algorithms in the estimation of user’s well-being for users that live alone (single-resident) or with others (multi-resident). Single-house and multi-house procedures are tested, both to extract a user-specific behaviour, and assess whether the model is able to generalise across different users and environments. Results show that the RF algorithm provides better performance than the RT algorithm in predicting the level of well-being with a Mean Absolute Error in the multi-house procedure of 32%, 13% and 17% for the Avg, Mind and Phy indices, respectively.

The rapid worldwide aging makes necessary the development of advanced technologies for the objective identification of diseases and disabilities onset. Accordingly, we need to regard the biological age as an alternative and more reliable indicator of the physiological decline of individuals, respect to the simple chronological age. In this paper, we present an eHealth system for estimate the biological age of elderly people starting from the assessment of the frailty phenotype. The frailty phenotype evaluation is made possible using a standard protocol for data acquisition and a cloud application for processing and storing data. The proposed eHealth system is also equipped with wireless, small and non-invasive wearable sensors for an objective evaluation of the mobility of a subject. The eHealth system is tested on a reference population in order to have an amount of data necessary for defining a model to estimate the biological age. The use of the presented system on a reference population, and the availability of data regarding their mobility, allow the validation of the proposed model for the computation of the biological age via simple and objective frailty phenotype assessment.

Measuring the behaviour and health status of informal caregivers of people with dementia can predict the personal well-being of the caregivers. Informal caregivers struggle to remain active during the daily life activities avoiding the care burden. For this reason, in this work, an analysis of both the environmental data coming from PIR sensors, installed in the home environment, and physiological parameters, directly measured by the user, is performed to highlight, unusual behaviors that can increase the stress level of the caregiver. In addition, daily survey and personal interview provide further information about the progression of the illness and the amount of the care burden. Coupling this information with the physiological quantities can provide an overall health status of the caregiver. Results show that the caregiver presents a decreasing trend of her daily self-reported health status associated with a change in the pattern of the domotic data. The questionnaire also exhibits a high correlation with body weight measurements (Pearson Coefficient of −86%) suggesting that the caregiver health status is limiting the normal daily activities, may be due to an increase of the care burden associated to a worsening of the illness.

Background: Brain-Computer Interface (BCI) is becoming more reliable, thanks to the advantages of Artificial Intelligence (AI). Recently, hybrid Deep Learning (hDL), which combines different DL algorithms, has gained momentum over the past five years. In this work, we proposed a review on hDL-based BCI starting from the seminal studies in 2015. Objectives: We have reviewed 47 papers that apply hDL to the BCI system published between 2015 and 2020 extracting trends and highlighting relevant aspects to the topic. Methods: We have queried four scientific search engines (Google Scholar, PubMed, IEEE Xplore and Elsevier Science Direct) and different data items were extracted from each paper such as the database used, kind of application, online/offline training, tasks used for the BCI, pre-processing methodology adopted, type of normalization used, which kind of features were extracted, type of DL architecture used, number of layers implemented and which optimization approach were used as well. All these items were then investigated one by one to uncover trends. Results: Our investigation reveals that Electroencephalography (EEG) has been the most used technique. Interestingly, despite the lower Signal-to-Noise Ratio (SNR) of the EEG data that makes pre-processing of that data mandatory, we have found that the pre-processing has only been used in 21.28% of the cases by showing that hDL seems to be able to overcome this intrinsic drawback of the EEG data. Temporal-features seem to be the most effective with 93.94% accuracy, while spatial-temporal features are the most used with 33.33% of the cases investigated. The most used architecture has been Convolutional Neural Network-Recurrent Neural Network CNN-RNN with 47% of the cases. Moreover, half of the studies have used a low number of layers to achieve a good compromise between the complexity of the network and computational efficiency. Significance: To give useful information to the scientific community, we make our summary table of hDL-based BCI papers available and invite the community to published work to contribute to it directly. We have indicated a list of open challenges, emphasizing the need to use neuroimaging techniques other than EEG, such as functional Near-Infrared Spectroscopy (fNIRS), deeper investigate the advantages and disadvantages of using pre-processing and the relationship with the accuracy obtained. To implement new combinations of architectures, such as RNN-based and Deep Belief Network DBN-based, it is necessary to better explore the frequency and temporal-frequency features of the data at hand.

In this paper we propose a fault/failure tolerant control scheme for a catamaran with two azimuth thrusters. The scheme is based on a three level architecture: a heuristic-based control policy for proper reference generation, a control law for the vehicle dynamics to achieve speed tracking of the generated reference, and a control allocation level for optimally redistributing the control efforts among the thrusters even in presence of actuator faults and failures; finally, the low level control is invariant with respect to the fault/failure scenario, as well as the control law. The proposed architecture has been simulated considering the nonlinear model of a Wave Adaptive Modular Vessel catamaran, with the aim of tracking a reference expressed in waypoints, even when faults/failures affect one or more thrusters.

Condition-based monitoring of rotating machines requires robust features for accurate fault diagnosis, which is indeed directly linked to the quality of the features extracted from the signals. This is especially true for vibration data, whose quasi-stationary nature implies that the quality of frequency domain extracted features depends on the signal-to-noise ratio (SNR) condition, operating condition variations, and data segmentation. This paper presents a novel statistical spectral analysis, which leads to highly robust fault diagnosis with poor SNR conditions, different time-window segmentation, and different operating conditions. The amplitudes of spectral contents of the quasi-stationary time vibration signals are sorted and transformed into statistical spectral images. The sort operation leads to the knowledge of the empirical cumulative distribution function (ECDF) of the amplitudes of each frequency band. The ECDF provides a robust statistical information of the distribution of the amplitude under different SNR and operating conditions. Statistical metrics have been adopted for fault classification, by using the ECDFs obtained from the spectral images as fault features. By applying simple statistical metrics, it is possible to achieve fault diagnosis without classifier training, saving both time and computational costs. The proposed algorithm has been tested using a vibration data benchmark: comparison with state-of-the-art fault diagnosis algorithms shows promising results.

Here, the authors present a low cost telerehabilitation system made up of a commercial red-green-blue depth (RGB-D) camera and a web-based platform. The authors goal is to monitor and assess subject movement providing acceptable and usable at-home remote rehabilitation services without the presence of a clinician. Clinical goals, defined by physiotherapists, are firstly translated into motion analysis features. A Takagi Sugeno fuzzy inference system (FIS) is then proposed to evaluate and combine these features into scores. In this stage, the `collaborative design' paradigm is used in depth and complete manner: the contribution of the clinician is not limited only to the rules definition but enters in the core of the evaluation algorithm through the definition of the fuzzy rules. A case study on low back pain rehabilitation involving 40 subjects, 5 exercises, and 4 physiotherapists is then presented to the effectiveness of the proposed system. Results of the validation of the system aimed at the assessment of the reliability of the proposed approach show high correlations between clinician evaluation and FIS scores. In this scenario, due to the high correlation, each FIS could represent a virtual alter-ego of the physiotherapist which enable a real time and free second opinion.

In a reconfigurable manufacturing scenario, control system design needs innovative approaches to face the rapid changes in hardware and software modules. The control system should be able to automatically tune its parameters to enhance machine performances and dynamically adapt to different control objectives (e.g., minimize control efforts or maximize tracking performances) while preserving at the same time stability and robustness properties. In this paper, a robust control system for permanent magnet synchronous motors (PMSMs), together with an online self-tuning method, is presented. In particular, a robust discrete-time variable structure control (VSC) has been designed. A heuristic bio-inspired approach has been then implemented on a digital signal processor (DSP) to find the VSC parameter set which minimizes a specific objective function each time a novel speed reference is provided. Experimental results on a PMSM motor show the effectiveness of the proposed controller and tuning method, with noticeable improvements with respect to the original manufacturer-designed controller.

This paper presents a method for the detection of series arc faults in electrical circuits, which has been developed starting from the recurrence quantification plots that allow to quantify the periodic behavior of time-series and to analyze the recurrences of a dynamical system presented by its phase space trajectory. Starting from this, the authors have found that it is possible to exploit recurrence quantification plots by using the gray-level co-occurrence matrix from which the extracted textural image features represent a proper set of indicators for suitably detecting arc faults. The database of this research is collected from 13 different types of load according to IEC 62606 standard. The proposed method’s effectiveness is shown by means of experimental tests, which were carried out in both arcing and non-arcing conditions and in the presence of different loads.

In this paper, a Hidden Semi-Markov Model (HSMM) based approach is proposed to evaluate and monitor body motion during a rehabilitation training program. The approach extracts clinically relevant motion features from skeleton joint trajectories, acquired by the RGB-D camera, and provides a score for the subject's performance. The approach combines different aspects of rule and template based methods. The features have been defined by clinicians as exercise descriptors and are then assessed by a HSMM, trained upon an exemplar motion sequence. The reliability of the proposed approach is studied by evaluating its correlation with both a clinical assessment and a Dynamic Time Warping (DTW) algorithm, while healthy and neurological disabled people performed physical exercises. With respect to the discrimination between healthy and pathological conditions, the HSMM based method correlates better with the physician's score than DTW. The study supports the use of HSMMs to assess motor performance providing a quantitative feedback to physiotherapist and patients. This result is particularly appropriate and useful for a remote assessment in the home.

Mobile and pervasive computing enabled a new realm of possibilities into the indoor positioning domain. Although many candidate technologies have been proposed, no one can still adapt to every use case. A case centered design and the implementation of the solution within the specific domain is the current research trend. With the rise of Bluetooth Low Energy (BLE) Beacons, i.e., platforms used to interact digitally with the real world, more standard positioning solutions are emerging in different contexts. However the reachable positioning accuracy with this technology is still unacceptable for some real applications (e.g., in the healthcare sector or the emergency management). In this paper, an hybrid localization application coupling a real time model based Pedestrian Dead Reckoning (PDR) technique and the analysis of the Received Signal Strength Indicator (RSSI) of BLE beacons is proposed. In particular, the smartphone application is composed by three main real time threads: a model based step length estimation, heading determination and the fusion of beacon information to reset the position and the drift error of the PDR. In order to give soundness to our approach we firstly validated the step length smartphone app with a stereo-photogrammetric system. The whole proposed solution was then tested on fifteen healthy subjects.

The joint position constraints of two cooperative robotic manipulators is a critical issue, especially when both manipulators grasp and move the same object. When two robots cooperate to move a grasped object from one point to another one, following a desired path, the general solution adopted in the literature is to use the kinematic redundant motions possessed by a single manipulator, in order to avoid joint constraints. However, this method requires a number of redundant motions at least equals to the number of joint constraints to satisfy. The study proposed in this paper is focused on developing a joint position limits avoidance strategy, which is able to satisfy all joint limits even when the number of redundant motions are no longer sufficient to ensure them with a classical approach. This is achieved by means of a supervisory control system, which allows to locally and temporary change the desired end-effectors motion, when the redundant motions are not available. The supervisory control sacrifices the path following task in order to ensure joint position limits avoidance, while preserving at the same time the relative end-effector motion, such that the stability of the grasped object is still ensured. The proposed strategy has been implemented on a commercial robot, namely Baxter, which possesses two anthropomorphic arms having one degree of redundancy. The obtained results prove the effectiveness of the proposed supervisor controller, which can be easily extended and applied in manufacturing scenario. Whereas path following can be temporary sacrificed (e.g., the grasped object can be moved from one point to another in different ways), the joint constraints must be strictly preserved for safety reasons.

The pillars of Industry 4.0 require the integration of a modern smart factory, data storage in the Cloud, access to the Cloud for data analytics, and information sharing at the software level for simulation and hardware-in-the-loop (HIL) capabilities. The resulting cyber-physical system (CPS) is often termed the cyber-physical manufacturing system, and it has become crucial to cope with this increased system complexity and to attain the desired performances. However, since a great number of old production systems are based on monolithic architectures with limited external communication ports and reduced local computational capabilities, it is difficult to ensure such production lines are compliant with the Industry 4.0 pillars. A wireless sensor network is one solution for the smart connection of a production line to a CPS elaborating data through cloud computing. The scope of this research work lies in developing a modular software architecture based on the open service gateway initiative framework, which is able to seamlessly integrate both hardware and software wireless sensors, send data into the Cloud for further data analysis and enable both HIL and cloud computing capabilities. The CPS architecture was initially tested using HIL tools before it was deployed within a real manufacturing line for data collection and analysis over a period of two months.

Nowadays identifying techniques aimed at a rational use of electric power has become even more important than the production of energy itself. This is caused by different factors, as the progressive saturation of the electricity grid, which is increasingly subject to connection requests, mainly due to the development of plants which exploit renewable energy sources. This work suggests a new approach based on the combination of the optimizer and the simulator developed in the MATLAB/Simulink environment, in order to reduce the energy costs in buildings during the summer while taking into consideration the user comfort. The electrical consumption of the entire building is taken into consideration is here examined with the aim of applying an air-conditioning system. The goal is to find, the day before, which is the optimal hourly scheduling of the control variables that must be applied the next day, taking into consideration all external conditions; weather conditions and the hourly energy price. In order to achieve this objective, the control variables, that have been changed, are the room temperature set points and the flow water temperature set point. As required by the UNI EN ISO 7730:2006 standard, comfort measurement is calculated by PPD (Predicted Percentage of Dissatisfied) index. Different scenarios are investigated and two optimization algorithms are compared. The results show that there is an average of 10 − 28 % potential cost saving, while maintaining a high level of comfort (PPD ≤ 12). The study is carried out by simulating a real office building in Italy, and the comparisons are shown regarding the actual settings applied to it.

Today, the improvement of the product value in consumer goods, such as new services to increase the positive customer experience, is the subject of many research activities. In a context where the product complexity becomes ever greater and the product life-cycle is always shorter, the use of intelligent tools for supporting all phases of the product life-cycle is very important. One of the aspects that is taking interest is to support the consumer in fault management. This analysis are well-known practices in the industrial, automotive fields, etc. but less used for consumer electronics. This paper analizes a Cloud service based on a Machine Learning (ML) approach used to provide fault detection capabilities to household appliances equipped with electric motors and compare the results with on premise ML algorithms provided research tools. The purpose of this paper is to perform a preliminary comparison of ML algorithm performances provided by two software, namely Microsoft Azure (cloud solution) and MATLAB (on premise solution), on a study case. In detail, the vibration data of an asynchronous motor installed in an oven extractor hood for commercial restaurant kitchen have been analyzed. To this end, two classification algorithms have been selected to implement fault diagnosis techniques.

Industry 4.0 is reinventing the way in which production is performed. Based on its eight pillars, I4.0 environments are adopting digital solutions in order to make production smart. One of these is the concept of augmented operator, which can act with the aid of digital tools to facilitate daily work. Augmented Reality can represent the turnkey. In this light, the aim of this research is to present a case study of a “security and safety” application through the use of AR smart glasses, tested in a real scenario. For our experiments, Vuzix Blade smart glasses have been tested in combination with a cloud-based architecture connected with an oil-extractor plant. The goal is to develop an AR application that allows to assist the operator during the working process. In particular, it acts as a guide system for the operator who wears glasses, provides remote support (remote operator) and, from a security point of view, sends real-time alerts in dangerous situations. The application has been validated after a number of practical tests carried out by specialised technicians who normally perform the work.

The fourth industrial age takes the manufacturing factory to a new level by introducing smart, extendible, flexible, modular and customized mass production technologies. Production lines or machines need to be integrated at the management level to be industry 4.0 compliant: in this way they can create and optimize a customer-oriented production, while constantly maintaining good performance conditions. In this context, one of the main challenges is the possibility to detect faults as fast as possible, to accurately diagnose those faults which can negatively affect the overall production cycle, and finally address them before it is too late. Due to the great importance that electric motors play in this context, an online smart algorithm for fault detection in electric motors is proposed in this paper. The effectiveness of the proposed method has been validated by applying it on an experimental benchmark, where the results show that the method is accurate and fast in detection of faults.

The identification of techniques aimed at a rational use of electric power has nowadays become more important than the production of energy itself. One of the causes for this is the progressive saturation of the Italian electricity grid, which is increasingly subject to connection requests, mainly due to the development of plants which make use of renewable energy sources.

The paper describes an approach for analyzing the use of a Laser-Guided Vehicle (LGV) in the context of the small and medium-sized enterprise. The use of LGVs is an efficient solution to provide more flexibility in the context of Just-In-Time production; however, the investment cost can limit this application. A methodology has been proposed in this work to analyze the technical feasibility of using an LGV in the manufacturing industry of customized products. The test case focuses on the study of a laser-guided system to optimize the handling of molds for customized production. In this scenario, an LGV is proposed to substitute manual carts used for moving molds from the warehouse to the injection machines. The traditional path included an intermediate station for pre-heating the molds in hot-air ovens. The proposed solution includes the study of an induction heating system on the LGV to optimize time and energy consumption.

Recurrence quantification analysis (RQA) allows to quantify the periodic behavior using recurrence plots instead of deriving information purely from visual analysis. The current study presents a preliminary analysis of stator-current measurements for electric motor fault detection and classification by means of the recurrence quantification theory. Firstly, a preliminary visual inspection of the recurrence plots of stator-current measurements for healthy and faulty electric motors is presented. Thereafter, the following RQ metrics are analyzed: the recurrence rate, the determinism, the divergence, the Shannon entropy, the laminarity and the trapping time. Then, the RQ metrics are used as predictors for fault detection and classification. The classification results (100% fault classification accuracy), which are presented using the linear support vector machine classifier, show that the RQA can be considered as a tool for motor current signature analysis.

In this paper, we propose an algorithm for snoring sounds detection based on convolutional recurrent neural networks (CRNN). The log Mel energy spectrum of the audio signal is extracted from overnight recordings and is used as input to the CRNN with the aim to detect the precise onset and offset time of the sound events. The dataset used in the experiments is highly imbalanced toward the non-snore class. A data augmentation technique is introduced, that consists in generating new snore examples by simulating the target acoustic scenario. The application of CRNN with the acoustic data augmentation constitutes the main contribution of the work in the snore detection scenario. The performance of the algorithm has been assessed on the A3-Snore corpus, a dataset which consists of more than seven hours of recordings of two snorers and consistent environmental noise. Experimental results, expressed in terms of Average Precision (AP), show that the combination of CRNN and data augmentation in the raw data domain is effective, obtaining an AP up to 94.92%, giving superior results within the related literature.

In this paper, a Machine-Learning Based Emotion Recognition System in patients with Parkinson’s disease is presented. The development of this system is composed of three steps. Firstly, each user is required to execute an experimental protocol while a simple device (i.e., smartwatch), worn on the wrist, collects data. During the experimental protocol, a nine-point clinical scale and a commercial emotion recognition software have been used to identify emotions. Secondly, from smartwatch data, features extraction is implemented. Lastly, a Machine Learning Algorithm (MLA) is trained with extracted features as input and emotion classes as output.

A kinematic controller for a dual-arm system able to cope with kinematic constraints is presented in this paper. The kinematic controller is designed according to the Relative Jacobian method to achieve cooperation of a couple of 7 DOF robotic arms. The kinematic redundancy obtained from cooperation is exploited to execute other subtasks along the main task. The concept of virtual joint space proposed in the General Weighted Least Norm (GWLN) method is employed in order to include the constraints in the problem. Firstly, the GWLN is reformulated for a dual-arm system, including only the joint limit avoidance subtask. Then, the obstacle avoidance subtask is also considered and a new version of the kinematic controller is derived when the number of constraints is larger than the number of joints. Simulation are performed on the model of the Baxter Robot, in the Matlab-Simulink environment.

This paper proposes a free dataset, available at the following link, 1 named KIMORE, regarding different rehabilitation exercises collected by a RGB-D sensor. Three data inputs including RGB, depth videos, and skeleton joint positions were recorded during five physical exercises, specific for low back pain and accurately selected by physicians. For each exercise, the dataset also provides a set of features, specifically defined by the physicians, and relevant to describe its scope. These features, validated with respect to a stereophotogrammetric system, can be analyzed to compute a score for the subject’s performance. The dataset also contains an evaluation of the same performance provided by the clinicians, through a clinical questionnaire. The impact of KIMORE has been analyzed by comparing the output obtained by an example of rule and template-based approaches and the clinical score. The dataset presented is intended to be used as a benchmark for human movement assessment in a rehabilitation scenario in order to test the effectiveness and the reliability of different computational approaches. Unlike other existing datasets, the KIMORE merges a large heterogeneous population of 78 subjects, divided into 2 groups with 44 healthy subjects and 34 with motor dysfunctions. It provides the most clinically-relevant features and the clinical score for each exercise. 1 https://univpm-my.sharepoint.com/:f:/g/personal/p008099_staff_univpm_it/EiwbKIzk6N9NoJQx4J8aubIBx0o7tIa1XwclWp1NmRkA-w?e=F3jtBk

The estimation of Biological Age (BA) has been debated for several years and no clear and universal understanding has yet been reached to solve this task. Accordingly, the knowledge of an accurate BA index for each individual may be relevant in various areas including health, economy, social policies and decision making processes. The main contribution of this work is the design of a Machine Learning based-consumer healthcare platform powered by electronic health record data (clinical features) and smartphone data (lifestyle features) in order to estimate a sub-index that is strictly correlated with the BA. Preliminary results extracted from a representative subset of clinical and lifestyle features, highlight the potential of the proposed framework in order to estimate the health and physical status of each subject (in terms of the difference between the predicted Chronological Age and the real Chronological Age). Future work will be conducted to encapsulate more information and validate the predicted BA sub-index.

Nowadays, one of the keys for the success of ICT technologies, in particular in the e-Health realm, is “Usability”. Self-assessment questionnaires are the most diffuse procedure for its evaluation. However, this procedure presents various limitations (e.g., subjectivity, needs of expert supervisors, etc.). In this paper, a methodology based on emotion recognition, web analytics and gaze detection is developed in order to asses usability of e-Health web platforms objectively, automatically, remotely and in the wild.

In this editorial, we provide an overview of the content of the Special Issue on “Smart Homes”. The aim of this Special Issue is to provide a comprehensive collection of some of the current state-of-the-art technologies in the context of smart homes, together with new advanced theoretical and technological solutions that enable smart technology diffusion into homes.

The human body has two different ages: a Chronological Age (CA), the actual time a person has been alive, and a Biological Age (BA), the real age that indicate the decline in health and in function ability during aging. Among previous studies, some authors proposed methodologies to estimate the Biological Age starting from non-invasive frailty measurements to evaluate the Frailty Index, others proposed invasive and expensive methods to measure the biological aging. Conversely, in this paper we propose a method to estimate the BA of a subject based on the assessment of the Frailty Phenotype. This type of evolution allows an efficient estimation of the frailty in contrast with the Frailty Index which is composed by a long checklist of clinical conditions and diseases to be evaluated from medical staff. We developed a cloud application, able to store and elaborate the data acquired during the evaluation protocol of the Frailty Phenotype, and also able to automatically provide the state of phenotypic fragility, and finally the Biological Age of a subject.

For several years, ENEA has been conducting a research project funded by the Ministry of Economic Development, within a broader context that envisages the "development of an integrated Smart Urban District model". Among the objectives of the project, there is the development of a replicable Smart Home model capable of stimulating energy savings and, at the same time, offering innovative services that aim to improve the quality of life of end users, in particular fragile users as well as older people. This paper presents the integration on the Smart Home developed architecture of a useful additional assisted living service, namely "MySignals" service, which is able to monitor the vital parameters of home users.

In this paper a Real-Time L-dopa-Induced Dyskinesia (LID) Detection System based on Machine Learning Algorithms (MLAs) and simple devices such as smartphone and smartwatch is presented. The implementation of this system was performed in three steps. Firstly, the data collection is carried out, where each patient wears the smartwatch and completes some tasks, while a smartphone application captures data. Secondly, features in time and frequency domain were extracted from smartwatch data and used as input for the training of different off-line MLAs. Lastly, the best algorithm found has been integrated into a mobile App in order to real-time monitor the smartwatch data and detect LID.

Nowadays the inspection of products at the end of line represents a critical phase. At this stage, it is necessary to look for defects in order to prevent the quality check to fail, and to provide information for improving the production as well. This task can be performed by using several sensing technologies, and the contactless optical inspection plays a key role. In this regard, the use of advanced robotic manipulators offers the capability to change the viewpoint of a given object and to inspect its multiple faces. We propose an approach that combines the use of photometric stereo to derive a 3D model of objects, empowered by the super-resolution that is applied on the original dataset (upstream) or on the normal images (downstream) in order to increase the quality of the final 3D model. The vision system is mounted on a robotic manipulator, able to grasp and change the viewpoint, thus offering a more complete view of the object to be inspected. The obtained results show that the developed solution increases the quality of the derived 3D models used for inspection tasks on different faces of the objects; this is achieved by using the manipulation ability offered by the adopted robotic platform.

The pillars of Industry 4.0 require a modern smart factory to be integrated, store data into the Cloud, access the Cloud for data analytics and share information at software level for simulation and Hardware-In-the-Loop capabilities. The resulting Cyber-Physical System is often called Cyber-Physical Manufacturing System, and it becomes fundamental to cope with the increased system complexity and the desired performances. However, since a lot of old production systems are based on monolitic architectures, with limited external communication ports and reduced local computational capabilities, it is very difficult to make such production lines compliant to Industry 4.0 pillars. Wireless Sensor Network is a solution for the smart connection of a production line to a Cyber-Physical System architecture, data processing through Cloud Computing. The scope of this research work is to propose an intermediate layer within the architecture that allows each device, production line and machine to be independently connected despite the adopted protocol. The solution is based on OSGi Framework, which is able to seamlessly integrate both hardware and software wireless sensors, send data into the Cloud for further data analysis, and grant both Hardware-In-the-Loop and Cloud Computing capabilities. A general description of the architecture is here proposed, together with preliminary results on a real manufacturing line for data collection and analysis over a period of two months.

This work has been developed within the framework of the Italian smart city project “Health@Home (H@H)”. The main goal is the development of a joint network (heterogeneous devices, both biomedical and home automation) to monitor the user’s health conditions within the home environment, together with dedicated services from the measured quantities. In particular, H@H follows the context of the Active and Assisted Living environment to improve the well-being of elderly giving support to the users at home. In this paper, the authors describe the implemented final prototype software architecture implemented, the measuring protocol used in resting conditions and the implementation of three user services, providing real-time feedback about the user health status.

Control Science has played over time and still continues to have a key role for the development of human society such that, in the XX century, it has been recognized as an independent discipline. During the ‘50’s, one of the first international scientific congress to address the area of control theory, namely, the “Convegno Internazionale sui Problemi dell’Automatismo”, was held in Italy, at the Museum of Science and Technology in Milan, Italy. In this context, ten years later, the Automatic Control group of the Universita di Ancona was born. In this chapter, we want to revisit the main results achieved in the last 50 years by the Automatic Control group of Ancona focusing, in particular, on the recent achievements about linear and nonlinear control; robust, stochastic, adaptive and optimal control; model predictive control; fault diagnosis and fault tolerant control; aerial, terrestrial and underwater unmanned vehicles control; automotive control; switching systems analysis and control; underwater robotics; intelligent manufacturing and cyber-physical production systems. The contribution aims to provide the main challenges on these topics and their future perspectives.

In this paper a L-dopa-Induced Dyskinesia Detection System based on Machine Learning Algorithms (MLAs), smartwatch data and a smartphone is presented. The development of this system was performed in three steps. In the first step each patient wears the smartwatch and fulfills some tasks while the smartphone Application captures data. The second phase is the features extraction from acceleration and angular velocity signals and the application of a Z-score normalization. In the last step two MLAs, trained with these features as input, are implemented in order to detect dyskinesias.

The Italian Ministry of Economic Development and the Italian National Agency for New Technologies, Energy and Sustainable Economic Development (ENEA) have entered into a Program Agreement for the execution of the research and development lines of General Interest for the national electricity system. In particular, as part of the “Development of an integrated model of the Urban Smart District” project, an experimental demonstration of a Smart Home network has been carried out in the Centocelle district of Rome, called “Centocelle Smart Home”. This project aims to develop a replicable model of Smart Home able to monitor energy consumption, the degree of comfort and safety in residential buildings, transmitting them to a higher level technological platform where data are analyzed and aggregated to provide a series of feedbacks to the user and to the community. The objective is to reduce the final consumption of energy (electric and thermal) of domestic consumers through a path of growth of energy awareness, also providing additional services on security. This paper presents the architecture proposed in the ENEA project for monitoring the energy consumption and increasing the degrees of comfort and safety of domestic users. The proposed security service is presented here together with the developed “Out of Home” App.

In this paper, we introduce a novel approach to design a user interface for commercial Smart Home Systems (SHS), following the needs of visually or hearing impaired users. The interface is able to transform visual information and alarms into audio signals and vice versa by using a mobile application. The aim of the interface is to make a commercial SHS usable for visually or hearing-impaired people, while maintaining a high level of acceptability, due to the use of an inclusive device, i.e., the smartphone.

Ahstract─ This work proposes a robotic manipulator assistant for disabled users and/or for elderly people with limited motor skills. In detail, the interaction among the robot and the user is based on the user gestures recognition. The user chooses an object among those available by moving his/her arm in a specific pose, which is recognized by using an external camera. Then, images of the objects accessible to the robot are acquired via the robot camera, located at the end of the robot arm, and are analyzed by a Support Vector Machine classifier in order to recognize the user selected object. Finally, the manipulator picks the object and places it on the user's hand, whose location in the Cartesian space is determined via the external camera and updated online.

In this paper a robust Fall Detection Algorithm by using a deep learning approach and a low-cost mobile robot equipped with an RGB camera is presented. This method consists of four steps. The first step is the user detection, achieved by a real-time video stream and a Deep Learning approach. Once the user is detected, then its position is estimated in the second step. In the third step, if a fall is detected, a photo is acquired and a pre-registered audio message asks the user how he is. In the last step the photo and the audio captured are sent to a Telegram Bot (TB) in order to alert family members or caregivers. Tests have been performed in a real scenario.

The paper is devoted to a new nonlinear discontinuous control algorithm and its applications in underwater robotics. The proposed approach is based on the usage of a relay control signals and coupled switching surfaces in order to improve the characteristics of a system. The choice of nonlinear switching functions to approximate own phase trajectories of a system in a case of relay control inputs, makes possible to get fast transient processes and high energy effectiveness. The derived approach is proven by applying it for controlling a remotely operated vehicle with parameters from a real ROV. Simulation results show the effectiveness of the proposed approach.

The goal of this work is to develop a smart light-emitting diode lighting system for industrial and domestic use with several advantages over conventional systems, namely energy saving, high reliability, and visual comfort of interior lighting. This is achieved by integrating a smart control module and a fault diagnosis and prognosis module within a conventional lighting system. The first module controls the lighting level in an energy-efficient way, keeping a desired light level where it is needed while regulating it to a minimum where not required; this is achieved by fully exploiting fuzzy logic and proportional-integrative-derivative controllers. The second module performs fault diagnosis on the light-emitting diode system and predicts when light-emitting diode maintenance should be performed; this is achieved by employing both hardware redundancy and signal-based approaches. Interaction between the two modules permits maintenance of a desired level of light even in the case of failures on one...

Summary The paper is devoted to a new adaptive variable-structure system (VSS) control algorithm and its applications in underwater robotics. The proposed approach is based on the usage of sliding mode parameter (SMP) in order to have adaptation in VSS. It is demonstrated that the extreme values of SMPs are directly associated with high efficiency of transient processes in VSS. The essence of the presented adaptation algorithm consists on the adjustment of switching surface parameters such that the SMPs move from the assigned desirable value to the extreme values, with the overall effect to accelerate transients. The effectiveness of the derived approach is proven by applying it for controlling a real remotely operated vehicle. Copyright © 2015 John Wiley & Sons, Ltd.

In this research, a localization system for a particular Automated Guided Vehicle (AGV), namely a Laser Guided Vehicle (LGV), which takes advantage of RSSI data, is described. The vehicle, its sensors and the wireless sensor network are virtualized, in order to provide a software replica of that available in a real factory. A Co-Simulator, composed by two simulation environments, has been implemented: from one hand, Robot Operating System (ROS), the RVIZ and GAZEBO for modeling and simulating the virtual system, and from the other hand, COOJA for reproducing the Wireless Sensor Network (WSN) behavior and RSSI signals. The virtualized LGV in ROS, together with its virtual sensors set, is integrated with the WSN in COOJA, so that localization can be performed by sensor fusion of odometric and RSSI data according to a Kalman Filter approach. The Co-simulator can be used to deploy path-tracking policies and test them before actual implementation in a Cyber Physical System (CPS) scenario.

This paper proposes a system for recognizing human complex activities by using unobtrusive sensors such as smartphone, smartwatch and bluetooth beacons. The method encapsulates two classification stages. The former is composed of two parallel processes: the Main Activity Detection (MAD) and the Room Detection (RD). The latter implements the Complex Activity Detection (CAD) process by exploiting the outputs of the first stage and the accelerometer data of the smartwatch. The cascade classification approach that combines the room detection with the main/complex activities recognition task constitutes the novelty of the work. Preliminary results demonstrate the reliability of the system in terms of accuracy and macro-Fl score.

Smart homes play a strategic role for improving life quality of people, enabling to monitor people at home with numerous intelligent devices. Sensors can be installed to provide a continuous assistance without limiting the resident&rsquo, s daily routine, giving her/him greater comfort, well-being and safety. This paper is based on the development of domestic technological solutions to improve the life quality of citizens and monitor the users and the domestic environment, based on features extracted from the collected data. The proposed smart sensing architecture is based on an integrated sensor network to monitor the user and the environment to derive information about the user&rsquo, s behavior and her/his health status. The proposed platform includes biomedical, wearable, and unobtrusive sensors for monitoring user&rsquo, s physiological parameters and home automation sensors to obtain information about her/his environment. The sensor network stores the heterogeneous data both locally and remotely in Cloud, where machine learning algorithms and data mining strategies are used for user behavior identification, classification of user health conditions, classification of the smart home profile, and data analytics to implement services for the community. The proposed solution has been experimentally tested in a pilot study based on the development of both sensors and services for elderly users at home.

In an energy management (EM) perspective, an objective function considering the balance of energy and resources needs to be optimized. To reach the goal, in a residential scenario, a number of tasks (typically appliances) can be managed (e.g. shifted). This paper proposes an unobtrusive cascade classifier approach to autonomously identify the different appliances and their priorities (shiftable, non interruptible, continuous loads) in EM algorithms based on smart plugs measures.

This study is realized within the framework of the Health@Home Italian project. The focus of this paper is to provide a description of the experimentation in a pilot case in Veneto region (Italy). The integrated residential sensor network (composed by a mix of domotic equipment and biomedical devices), which will allow older people to improve their life-style in their houses, is discussed, together with the apartments selection (i.e. the specifications and requirements) and the users’ recruitment. The authors will also introduce the analysis of the expected results, based on measurements of preliminary data and signals in living lab and the first feedback from the 13 recruited users. The results of the preliminary tests are used to improve the architecture, following the user-acceptance and the data collection. In this phase of the study, the possible services have been hypothesized, and this aspect will be investigated after the end of the experimentation phase (end of 2018)

This paper presents a preliminary set of results on the design of a virtual thruster-based failure tolerant control scheme for underwater vehicles. The proposed control scheme is based on the use of a suitable thruster allocation algorithm, which consists on a modified version of the Moore-Penrose pseudo inverse. When a thruster experencies a failure, the “virtual” thruster concept is introduced, as an ideal faultless thruster that operates in places of the real failed one. Thus, the resulting thrust force, which should be allocated to the failed actuator (i.e., virtual thrust), is instead reallocated to the still functional thrusters. A bank of controllers is built so that each controller is designed to control the considered underwater vehicle under a specific actuator failure scenario.

This work proposes an approach for performing Human-Robot Cooperation (HRC) tasks by integrating a Brain Computer Interface (BCI) with a robotic manipulator. In detail, the user can select one among six different objects via BCI, which analyzes the P300 signals generated by the brain when the images of the selectable objects appear on the screen. Then, the selected object is recognized by a Support Vector Machine (SVM) classifier, which scans the data acquired by the robot camera. Finally, the recognized object is grasped by the robotic arm and moved to a different position.