Ubiquitous Computing and Ambient Intelligence

This contribution presents a proposal for generating linguistic reports based on the study of biomedical signals of human patients. Although this topic is dealt in many previous works, there are challenges still open for the scientific community, such as the development of systems to produce reports and alerts using a human-friendly language. We present a brief review of some relevant previous works, as well as our proposal of a system based on fuzzy linguistic approach applied to the diagnosis of the preeclampsia disease that may affect pregnant women. Our system transforms numerical values of biomedical signals into linguistic values that are understandable information for the patients and the medical staff. The dataset used for testing the system contains real data from a study carried out by the Davinci UNAD Group (Colombia) on patients that suffer from preeclampsia.

Falls in hospitals, in residential care facilities and in home of elderly commonly occur near the bed. Recognizing bedside events may give caretakers the opportunity to intervene, thereby preventing a fall from happening. Most approaches today either use cameras which invade privacy, or sensor devices attached to bed. In this paper an experimental approach for recognizing bedside events using a ceiling mounted 60 × 80 longwave infrared array combined with an ultrasonic sensor device is presented. This novel approach makes it possible to monitor activity while preserving privacy in a non-intrusive manner.

In this paper we propose a vision based gait analysis approach to work with frontal view sequences. The main issue of sagittal view gait sequences is the physical space required to record them. We propose two different approaches to obtain heel strike and toe off with frontal gait, both of them are based in the time series of the difference of component y of both feet. In the former, the zero crosses are used to determine the range in which heel strike and toe off occurs. In the latter, the maxima and minima are used instead. Testing our approach with our own dataset show that it is possible to obtain heel strike and toe off events using only frontal view gait sequences recorded with an RGB camera. Results show as well that it is possible to classify between normal and abnormal gait using frontal view.

Feature subset selection is an important data reduction technique. Effects of feature selection on classifier’s accuracy are extensively studied yet comprehensibility of the resultant model is given less attention. We show that a weak feature selection method may significantly increase the complexity of a classification model. We also proposed an extendable feature selection methodology based on our preliminary results. Insights from the study can be used for developing clinical decision support systems.

Knee alignment measurements are one of the most extended indicators of knee-complex injuries such as anterior cruciate ligament injury and patellofemoral pain syndrome. The Frontal Plane Projection Angle (FPPA) is widely used as a 2-D estimation of knee alignment. However, traditional procedures to measure this angle suffer from practical limitations, which leads to huge time investments when evaluating multiple subjects. This work presents a novel video analysis system aimed at supporting experts in the dynamic measurement of the FPPA in a cost-effective and easy way. The system employs Kinect V2 depth sensor to track reflective markers attached to the patient leg joints to provide an automatic estimation of the angle formed by the hip, knee and ankle joints. Information registered by the sensor is processed and managed by a computer application that simplifies expert’s work and expedites the analysis of the test results.

A methodology for mining data coming from mobile phone accelerometers is proposed in order to discover movement patterns in Alzheimer patients and to explore the relation of these patterns with the stage of the disease. This methodology processes the data provided by the accelerometer to extract features of the patient movement patterns. This information is used to train a neural network that relates the patient movement patterns with the stage of the disease (early, middle or late). This proposal based on neural network classifiers is compared with other machine learning classifiers. Moreover, this methodology is applied in a case study with 35 patients. Initial experiments are promising with a success rate up to 83 percent. The projection and exploitation of the results of our analysis are subject to ulterior extensive validation of the proposed technique.

Hypertension is a common and dangerous condition, which is the most important preventable cause of stroke and heart disease. Long-term conditions result in a reduced quality of life that can be improved through self-management and empowerment of patients using information technologies. Current support systems include self-management and empowerment in patients, but both features are not personalised in terms of patient preferences and decision-making. In this work an adaptive genetic algorithm is proposed for personalised support systems in hypertensive patients by including patient blood pressure data in the generational replacement step of evolutionary computing.

Crohn’s disease belongs to the group of inflammatory bowel diseases. The current process of disease management has significant weaknesses that cause high cost for health systems and a significant loss of quality of life for the patient. This paper shows a new approach to redesign process for the management of Crohn’s disease based on Business Process Management strategy. This approach seeks to improve the patient empowerment and self-management, reducing costs and obtaining constant and updated information throughout the process.

Chronic pelvic pain is a common clinical condition with negative consequences in many aspects of womens life. The clinical presentation is heterogeneous and the involvement of several body systems impairs the identification of the exact etiology of the problem. At the same time, a clinical treatment of good quality depends on the professional and the learning process is slow. The goal of the paper is to show techniques used to create an artificial intelligence system capable of indicating the probable causes of this condition in order to help the doctors in the diagnosing process. This system uses a supervised learning algorithm along with multi-label problem modeling techniques and attribute selection algorithms to achieve the desired goal.

Three-Dimensional (3D) technologies emerged from the technological advances in manufacturing required to produce physical versions of digital models. The most attractive feature of 3D technologies is that virtual models are easy to mold, and custom-made items can be physically produced. Health domains are areas in which 3D technologies have been applied, and several studies have been conducted assessing the usefulness of such technologies in those domains. In this paper we present the results of a Systematic Literature Review (SLR) on the applications of 3D technologies in the health domain. Discussion from the revision of 33 papers is presented. The main finding of this SLR is that none of the available research papers are focused on computer science related areas (i.e., all papers are published by doctors or researchers in Medicine). Moreover, all the included papers were published in journals specialized in Medicine. Therefore, they do not delve in the computational conclusions of the studies. In this article, we identified significant research gaps (from the computational perspective), as well as new ideas are being proposed on the future of 3D technologies in health.

The growing interest in computer assisted rehabilitation to alleviate the lack of enough facilities and specialists to cope with current demand for rehabilitation, especially related to the ageing of population, has pushed forward challenges and innovation related to the design and development of such systems. One of the aspects present in rehabilitation is motivation. Motivation is not essential in rehabilitation, but has been proven a useful factor to increase the efficiency of a rehabilitation process. In this paper we discuss the concept of motivation by providing a model that aims at supporting the design of the characteristics of motivation. Influence Awareness model is introduced as a vehicle to provide the patient with the information required to influence her behavior to improve her motivation towards rehabilitation in a computer assisted environment. Moreover, this Influence Awareness is integrated into a modelling language that enables the specification of the tasks to be accomplished during rehabilitation. Lastly, these concepts are exemplified in a case study.

In this paper we propose a vision based gait analysis approach that work under real time constraints. We propose the use of a multiresolution pyramid image representation that allows to provide suboptimal responses if the deadline is reached. The impact of each suboptimal response is analysed showing that although there is an impact in the quality of the output, the gait analysis algorithm still provides satisfactory results. In addition, the adjustment to time constraints of the proposed approach is also analysed showing suitability for real time constraints.

When collaborating remotely, being aware of other participants (their actions, locations, status, etc.) is paramount to achieve a proper collaboration. This issue is magnified when talking about rehabilitation systems, whose users may require additional specific awareness information, due to their cognitive or physical disabilities. Moreover, because of these disabilities, this awareness may be provided by using specific feedback stimuli. This constituted the main motivation of this work: the development of an awareness interpretation for collaborative cognitive and physical therapies. With this aim, an awareness interpretation already applied to the collaborative games field has been modified and extended to make it suitable for these systems. Furthermore, in order to put this interpretation into practice, a case study based on an association image-writing rehabilitation pattern is presented illustrating how this cognitive rehabilitation task has been extended with collaborative features and enriched with awareness information.

Diabetes is a chronic disease that requires continuous medical care and patient self-monitoring processes. The control of the glucose level in blood is a task that the patient needs to perform to prevent hypoglycemia episodes. Early detection of hypoglycemia is a very important element for preventing multi-organ failure. The incorporation of other biomedical parameters monitoring, combined with glucose levels can help to early detect and prevent those episodes. At this respect, several e-health platforms have been developed for monitoring and processing vital signals related to diabetes events. In this paper we evaluate a couple of these platforms and we introduce an algorithm to analyze the data of glucose, in order to anticipate the moment of an hypoglycemia episode. The proposed algorithm contemplates the information of several biomedical sensors, and it is based on the analysis of the gradient of the glucose curve, producing an estimation of the expected time to achieve a given threshold. Besides, the proposed algorithm allows to analyze the correlations of the monitored multi-signals information with diabetes related events. The algorithm was developed to be implemented on an FPGA-based SoC and was evaluated by simulation. The results obtained are very promising and can be scalable to further signals processing.

Hypertension affects eight out of ten adult population over 65 years. Healthcare processes require interdisciplinary cooperation and coordination between medical teams, clinical process and patients. The lack of patients’ empowerment and adherence to treatment makes it necessary to integrate patients, data collecting devices and the clinical process together. The use of Business Process Management (BPM) paradigm and their associated technologies as an integrating tool throughout the clinical process of coordinating the data collected hypertension patients through the devices, the clinical process and the needs of the medical equipment is proposed.

In recent years, wearable sensors are integrating frequently and rapidly into our daily life day by day. Such smart sensors have attracted a lot of interest due to their small sizes and reasonable computational power. For example, body worn sensors are widely used to monitor daily life activities and identify meaningful events. Hence, the capability to detect, adapt and respond to change performs a key role in various domains. A change in activities is signaled by a change in the data distribution within a time window. This change marks the start of a transition from an ongoing activity to a new one. In this paper, we evaluate the proposed algorithm’s scalability on identifying multiple changes in different user activities from real sensor data collected from various subjects. The Genetic algorithm (GA) is used to identify the optimal parameter set for Multivariate Exponentially Weighted Moving Average (MEWMA) approach to detect change points in sensor data. Results have been evaluated using a real dataset of 8 different activities for five different users with a high accuracy from 99.2 % to 99.95 % and G-means from 67.26 % to 83.20 %.

In this research work it is presented a preliminary prototype for an ambient intelligence scenario based on cyber-physical systems for improving learning tasks. The system proposed is composed of a cyber-glove, a worktable (both with RFID and NFC detection zones) and a AmI software application for modeling and workflow guidance. The authors carried out a case study where 12 mentally handicapped people and 3 trainers were involved executing workflows creation and performing and controlling tasks. The results obtained indicate that this kind of solutions are feasible, but due to the problem complexity and to the fact that the proposed solution is a preliminary version, we have found many issues to be solved in next versions. This research helped us to uncover these issues and design a better system.

This work is related to the development of a personalised machine learning algorithm that is able to classify food images for food logging. The algorithm would be personalised as it would allow users to decided what food items the model will be able to classify. This novel concept introduces the idea of promoting dietary monitoring through classifying food images for food logging by personalising a machine learning algorithm. The food image classification algorithm will be trained based on specific types of foods decided by the user (most popular foods, food types e.g. vegetarian). This would mean that the classification algorithm would not have to be trained using a wide variety of foods which may lead to low accuracy rate but only a small number of foods chosen by the user. To test the concept, a range of experiments were completed using 30 different food types. Each food category contained 100 images. To train a classification algorithm, features were extracted from each food type, features such as SURF, LAB colour features, SFTA, and Local Binary Patterns were used. A number of classification algorithms were used in these experiments; Nave Bayes, SMO, Neural Networks, and Random Forest. The highest accuracy achieved in this work was 69.43 % accuracy using Bag-of-Features (BoF) Colour, BoF-SURF, SFTA, and LBP using a Neural Network.

The paradigm of healthcare systems has change from isolated proprietary health records to patient-centric solutions in which government, hospitals and clinics, general practitioners and other stakeholders must cooperate in order to provide improved health services. Enabling interoperability to share heterogeneous medical and administrative information in a secure environment is an issue addressed worldwide. Standards help though are not enough to provide the right information at the right time and place. In this paper we proposed to leverage the interoperability between standards through the mediation of a ubiquitous user model and an automatic process of concept alignment.

In mobile sensing, modern phones allow researchers obtain the information about the participants and their surroundings in a precise, unobtrusive, unbiased, and timely way. However, obtaining this information is just the first step of the research work, concentrating, processing, and giving meaning to the collected data also require a considerable amount of effort. In this work, we present a platform that addresses the aforementioned activities by providing a means to obtain data through sensors of a mobile phone and process those data in the mobile phone, prior to sending them to an online repository.

Mobile sensing has become a growing area of research in pervasive healthcare. In this paper we present mk-sense, an open framework for mobile sensing on smartphones. mk-sense is an initiative to reduce the efforts of researchers involved in multi-institutional sensing campaign. It is designed to facilitate the collaboration of researchers that run simultaneous data collection efforts in different locations. We illustrate the use of mk-sense with two cross-cultural studies conducted in four different countries (Turkey, Mexico, Switzerland, and Spain) with a total of 77 participants. In this paper, we describe the challenges and experience of conducting research in the wild by using mk-sense as sensing platform. Finally, we present how the conducted studies influenced the design decisions of mk-sense, including features, and tools to monitor data gathering in real-time.

Recent advances in wireless sensors technology applied to e-health allow the development of “personal medicine” concept, whose main goal is to identify specific therapies that make safe and effective individualized treatment of patients based, for example, in health status remote monitoring. Also the existence of multiple sensor devices in Hospital Units like ICUs (Intensive Care Units) constitute a big source of data, increasing the volume of health information to be analyzed in order to detect or predict abnormal situations in patients. In order to process this huge volume of information it is necessary to use Big Data and IoT technologies. In this paper, we present a general approach for sensor’s information processing and analysis based on Big Data concepts and to describe the use of common tools and techniques for storing, filtering and processing data coming from sensors in an ICU using a distributed architecture based on cloud computing. The proposed system has been developed around Big Data paradigms using bio-signals sensors information and machine learning algorithms for prediction of outcomes.

In this paper, we present an Android application to control and monitor the physiological sensors from the Shimmer platform, which include ECG (Electrocardiogram), EMG (Electromyogram), and GSR (Galvanic Skin Response) modules and accelerometer, magnetometer, and gyroscope. The application can configure, select, receive, and represent graphically and store the signals from the sensors. Experimental results with two Android devices were carried out. The ECG, EMG, GSR, and gyroscope sensors were monitored simultaneously at a sampling rate of 10.2 Hz. The application can be applied to many different users such as patients with chronic diseases, athletes, or drivers.

Technologies to monitor patients are convenient for patients and can reduce health costs. Chronic pain is a pain that lasts more than 3 months and affects the welfare of patients. Pain is subjective and there are applications to self-report pain, but their adherence rates are low. The purpose of this article is the understanding of the characteristics of technology that helps the adoption of these systems. We have implemented two solutions (mobile application and wearable device), in order to compare them to measure the rate of user acceptance, and also to get feedback about fundamental features of interfaces to report pain levels. To evaluate the two solutions we conducted interviews with 12 people. The results showed that when given the choice between both devices, 67 % of the users preferred the wearable device over the mobile application, and 16.5 % preferred the mobile application over the wearable device. We also found that a device for reporting pain must be specific to this purpose, aesthetically pleasing and allow users to report easily and at the right time.

This project is focused on the design and development of a mobile application that records the daily food intake for a person, and the caloric based-content indicates that it is the amount of calories consumed. To perform the calculation of Basal Metabolic Rate (BMR) Harris Benedict equation is used, this method of calculation is based on age, sex, height and weight. The estimated value is multiplied by a number corresponding to the level of activity of the individual. The resulting number is the recommended kilo calories to maintain current body weight daily intake. The system will have two modules, one is that which manages the data capture and delivery for storage. This in turn consists of a mobile application with a simple user interface where you can enter data and prepares to send them to the storage service. In addition, the App can view the recorded data and display it on the screen data queries.

Workflow management has been shown to be a promising approach to the support of a range of healthcare processes, with tools available for their formal specification, analysis and implementation. To further illustrate its relevance, we apply a workflow modelling approach to the specification and analysis of an anticoagulation monitoring protocol, illustrating a Petri Net-based solution using YAWL and Coloured Petri Nets. The selected scenario is representative of healthcare processes which have not been extensively considered for workflow solutions in the past – namely highly distributed, multi-party activities executing over an extended period of time. In presenting a workflow analysis for such a case, we identify challenges in supporting these types of primary and community care-based processes and identify possible areas in which workflow solutions could be extended to address their particular process requirements.

Children with physical impairments, ranging from impaired mobility to very limited mobility, often require mobility aids to compensate for these difficulties. These impairments can adversely affect the child to varying degrees and have an impact on their health and wellbeing. It is estimated that 30 %–40 % of medical interventions have no reported evidence base and another 20 % of interventions delivered are ineffective. Clinicians are under increasing pressure to provide evidence of the effectiveness of prescribed treatments and products. Therefor there is a need to provide clinicians with empirical data that evidences practice and provides a quantified assessment of treatment efficacy through data gathering in both real-time and longitudinally, combined with data analytics to further develop treatment strategies. This paper presents a system to assist and enable clinicians to analyze and asses the effectiveness and usage of prescribed treatments for physically impaired children. The system achieves this through the use of a gamified data collection app and a web portal to analyze and present summarized measures of gait.

Appointment lead-time is a critical variable in outpatient clinic services. In Gynecology and Obstetrics departments, longer appointment lead times are associated with lower patient satisfaction, the use of more complex healthcare services, development of long-term and severe complications and the increase of fetal, infant and maternal mortality rates. This paper aims to define and evaluate improvement alternatives through the use of Discrete-event simulation (DES). First, input data analysis is performed. Second, the simulation model is created; then, performance metrics are calculated and analyzed. Finally, improvement scenarios are designed and assessed. A case study of a mixed-patient type environment (Perinatology and Gynecobstetrics) in an outpatient department has been explored to verify the effectiveness of the proposed approach. Statistical analysis evidence that appointment lead times could be significantly reduced in both Perinatology and Gynecobstetrics appointments based on the proposed approaches in this paper.

The United Nations Children’s Fund (UNICEF) is a UN organization whose charter is to protect and improve the lives of children around the world. Maternal and child health are health-related areas where UNICEF has developed innovative Information and Communication Technology (ICT) solutions in the general domain of mHealth in which text messages have been used to address particular health issues. We have used two UNICEF open source software packages, Rapid SMS and Rapid Pro, in tele-health projects in Nicaragua. In this paper we describe the implementation of these projects and the relative advantages/disadvantages of using these two software tools in implementing our solutions.

This paper presents the use of discrete-event simulation to support process improvements at an outpatient internal medicine department. This department is significantly effective upon treating patients; however, patient waiting times tend to be longer and consequently patient satisfaction rates continue to decrease. With the aid of this technique, 3 improvement scenarios proposed by medical and administrative staff from this department were designed and simulated including changes related to installed capacity and an emphasis on physicians keeping to the schedule. Statistical analysis of output data evidenced which scenarios resulted in poor performance (statistically equal or higher waiting times) and which strategies caused lower waiting times. In this case, Scenario 3 was selected as the best improvement choice with 71.28 % and 19.28 % reduction in average waiting time and standard deviation respectively. With this approach, inefficient strategies can be avoided and real improvement alternatives can be identified.

Clinical gait analysis provides an evaluation tool that allows clinicians to characterize person’s locomotion at a particular time. There are currently specialized systems to detect gait events and compute spatio-temporal parameters of human gait, which are accurate and redundant. These systems are expensive and are limited to controlled settings with gait evaluations widely spaced in terms of time. As alternative, a proposal for long-term gait monitoring in Assisted Living Environments based on an infrastructure of wireless inertial sensors is presented. Specifically, heel-strike events will be identified in multiple elders in a rest home and throughout the day. A small wearable device composed of a single inertial measurement unit will be placed at the back of each elder, on the thoracic zone, capturing trunk accelerations and orientations which will enable the demarcation of heel-strike events and the computation of temporal gait parameters. This proposal attempts to contribute to the development of a less intrusive and reachable alternative for long-term gait monitoring of multiple residents, which has been poorly investigated.

In this paper, it has been analysed the responses, in terms of cognitive activation through EEG, to specific external stimuli. It has been developed a videogame, as a particular kind of serious games from health, which promotes the exercise of cognitive abilities. The participants were ten healthy children between 7 to 12 years old, selected because of their developmental age. Mechanics included in the videogame have been evaluated and related to the processing of new information and user response in preliminary works. The hypothesis of the current work refers to how specific mechanics that are involved in platform-videogames cause activation in the electroencephalogram waves according to cognitive processes such as short-time memory, attention and concentration. It has been analyzed through the magnitude of EEG frequency bands. The results are consistent by showing a differential activation during several game mechanics. With these results we can conclude that the videogame promotes activation and exercise in areas related to the mentioned cognitive skills when the participants are playing the videogame mainly during particular mechanics.

This paper investigates the user experience design of a smartphone application for the objective assessment of cognitive fatigue. This is as an alternative to using an established paper questionnaire that offers subjective self-assessment. Taking a multidisciplinary approach, challenges relating to the usability and the efficacy of the smartphone assessment tool were explored. Furthermore, to enable validation of the proposed new approach, challenges relating to how best to deliver the traditionally paper-based questionnaire on a smartphone display, while retaining the validity of the measure it affords, had to be addressed. Results show that the smartphone based cognitive testing methods was comparable to outcomes from the pre validated mobile based Mental Fatigue Scale. Participant feedback showed that the smartphone-based approach offered a more acceptable and engaging user experience, while retaining the ability to accurately measure cognitive fatigue.

This paper provides an overview of a doctoral research project which focuses in developing a framework to allow smart device-based notifications to provide user awareness. Notifications are mechanisms by which the user’s attention is driven to specific tasks or events. Notifications should provide a balance between intrusiveness and value in order to avoid annoyed users. The results reported in this paper include 50 % of the overall results expected for the project. The first step was to conduct a systematic literature review assessing the use of smart devices to deliver notifications, the second step was the development of a framework to allow notification coordination among smart devices and the third result is a laboratory case study assessing the framework. Future work includes two case studies in real scenarios and further analysis of usage patterns identified during this research.

An important factor for the next generation of Human Computer Interaction is the implementation of an interaction model that automatically reasons in context of the users goals, attitudes, affective characteristics and capabilities, and adapts the system accordingly. Although various techniques have been proposed for automatically detecting affective states using facial expression, this is still a research challenge in terms of classification accuracy. This paper investigates an extensible automatic affective state detection approach via the analysis of facial expressions from digital photographs. The main contribution of this study can be summarised in two points. Firstly, utilising facial point distance vectors within the representation of facial expressions is shown to be more accurate and robust in comparison to using standard Cartesian coordinates. Secondly, employing a two-stage Support Vector Machine-based classification model, entitled Hierarchical Parallelised Binary Support Vector Machines (HPBSVM), is shown to improve classification performance over other machine learning techniques. The resulting classification model has been evaluated using two different facial expression datasets (namely CKPLUS and KDEF), yielding accuracy rates of 96.9 % and 96.2 % over each dataset respectively.

In this paper, we propose a new design strategy for integrating fictionality into the real world named Alternative Reality, which makes it possible to connect the daily urban world with the virtual world from a temporal aspect to influence humans to adopt better lifestyles. The worlds also can be seamlessly integrated because the virtual world consists of real landscapes, objects and persons. This means that it may be possible to enhance the real world by showing fictional events among real events: people experience the enhanced hybrid world as in the real world rather than in a fictional world such as a movie. To demonstrate the design strategy of Alternative Reality, we have developed two case studies. The first case study investigates whether a user can sense the improbable behavior of a moving object as realistic, where the user can interact with the object. The second case study investigates whether a user can experience fictional occurrences in the virtual world as they are experienced in the real world. In both case studies, a user wears a head-mounted display to increase the immersion in the hybrid world created by Alternative Reality, in which the virtual world is created by capturing the real world with a 360-degree camera. The insights of the experiments with the case studies show that Alternative Reality effectively augments the real world without losing touch with reality.

Nowadays, the rapid spread of new technologies has certainly revolutionized the way how people communicate and access information. Cultural practitioners deal with this issue endlessly and look to the new technologies as an opportunity to enhance their results in terms of exhibit experience and expressiveness. Such technology adoption needs to come along with the development of tools that make it possible for cultural practitioners to freely develop their ideas. The design of such tools for a specific category of users requires a deep knowledge of that category in order to establish appropriate usability requirements. In this paper we present a participatory design process aimed at defining solutions to support cultural practitioners in developing augmented reality applications for the specific cultural domain.

The use of smart devices is increasing rapidly; this trend is changing the paradigm in which notifications are delivered to users. Smart devices are important to provide user awareness. However, their use must be controlled and human perception should be considered to avoid information overload. In this paper, we present a dynamic mechanism to coordinate the distribution of notification across smart devices. This personalized notification mechanism uses an inference engine and a set of rules to generate notification alternatives and select the “best” one. A continuous refinement approach is also used to improve notification delivery. Our system was evaluated and the baseline rules were established by 11 expert users. The main results show that in some scenarios, the notification mechanism selection converged quickly and results are promising. However, further work is required to provide not only personalized but integrated (i.e., more than one device at the time) notification management.

The Smart Cities discourse intends to envision the future of cities. It is mostly orientated towards an efficient and optimal management of citys resources to enhance peoples life. Our relation with the city environment is changing since connected objects are incorporated in the city bridging the physical with the virtual world. With new opportunities of engaging citizens in new relations within the city, what we present is a multidisciplinary framework for observing those interactions. It combines different methods from Computer Sciences, Social Sciences and Design, to evaluate the assemblages in the urban space with the idea to understand the benefits of affective and empathic relations in long-term interactions with software enabled objects. While this framework will be tested in different experiments, the ultimate aim is to provide knowledge and insightful support for designing future objects for the urban space that enhance the experience of people’s everyday life.

Students’ participation in traditional classroom settings may be hindered due to various reasons, which interrupt the class flow or cause distraction among the rest of the class members. To tackle that problem, we propose using applications based on touchless hand gestures (THG) that would allow students to interact from their own places. The feasibility of this proposal is analyzed in this work. To do it, we requested students to use two applications from their physical locations. Obtained qualitative results suggest the proposal may be used in an acceptable way as the use of applications based on THG becomes widespread. Actually, students who participated recommend the use of this proposal, since they may be more motivated to participate actively in the development of classes, which would result in better teaching-learning processes.

People with disabilities have fewer opportunities. Technological developments should be used to help these people to have more opportunities. In this paper we present partial results of a research project which aims to help people with disabilities, specifically deaf and hard of hearing. We present a sign language recognition model. The model takes advantage of the natural user interfaces (NUI) and a classification algorithm (support vector machines). Moreover, we combine handshapes (signs) and non-manual markers (associated to emotions and face gestures) in the recognition process to enhance the sign language expressivity recognition. Additionally, non-manual markers representation is proposed. A model evaluation is also reported.

A prominent requirement in the field of human-computer interaction is to make mobile applications more usable and better adjusted to their users’ needs. In particular, designers of groupware applications face the task of developing software for many users while making it work as if it was designed for each single individual. User modeling research has attempted to address these issues. A precondition for achieving this task is to find predictive and generative models of the user interactions. In this paper we develop a methodology for modeling the user behavior when interacting with a computer system. The byproduct of this methodology is a low level representation of the user interactions in the form of weighted automata, which can be easily transformed into user profiles in text form. Profiles can then be used by the designer to configure and verify the task model of the system.

This paper presents an experimental study to assess the capabilities of older adults to interact with multi-touch surfaces. The study involved 100 elderly people between 61–92 years old. We selected two different elderly centres in Madrid, with different characteristics in terms of income level. The “Gesture Games” tool was used because it allows experimenting with the seven more used multi-touch gestures: Tap, Double tap, Long press, Drag, Scale up, Scale down and One-finger rotation. The analysis of the data showed that older adults have total capacity to execute these seven tasks. Some of the tasks, such as “scale down” and “scale up” were found easier for them, while other tasks, such as “double tap” were more difficult.

This paper presents a proposal for integrating people in risk of exclusion by means of Virtual Worlds. Different studies show the benefits of using virtual worlds for educational purposes, where you can develop a wide variety of innovative teaching and learning activities. In particular we are working with immigrants; with high school students with learning disabilities; and with people with cognitive disabilities (autism, Asperger, Down syndrome). In this paper we present three experiences in this area, one of them already finished and the other two being currently implemented.

Self-adaptation is a key requirement in emerging software systems that must become capable of continuously adapting its behavior at run-time to their context (new environmental conditions, resource variability, unpredictable situations, changing user needs, etc.) without human intervention. However, experience in autonomous systems shows that people cannot be excluded entirely of the adaptation loop. For example, in the case of autonomous cars, they still need humans to drive in certain situations (e.g., complex driving situations, emergencies, etc.). This work defines the key factors to design the human participation in the control loops by introducing a framework to design human participations. Our framework considers human attention as a critical factor for user participation. Also, it pays attention to the dynamism between different types of human participation depending on the different system limitations (e.g., uncertainties in sensing, conflicts in goals, etc.) and the current user situation (e.g., user attention, environmental situation, etc.). We illustrate our approach by applying it to manage some actual autonomous cars situations that require human intervention.

We present an immersive virtual reality tool, called VR GREP, to empower end users with the capacity to design and develop virtual reality environments by themselves. To investigate the potential benefits that this technology might provide to support the end user development of virtual environments we conducted a study in which 23 participants collaborated. The participants designed and implemented two virtual environments using their favourite interaction style. They reported that the immersive environment contributed to making them feel more creative and engaged in the process. They also agreed that the perspective provided by the tool was more adequate for the process, as it is closer to the one perceived by the final user. In general, the results suggest that the technology could have a great potential for supporting the authoring tasks, although to make it a viable alternative to desktop based solutions the precision and accuracy when interacting with the virtual environment will need to be improved.

Telepresence robots have emerged as a new means of interaction in remote environments. However, the use of such robots is still limited due to safety and usability issues when operating in human-like environments. This work addresses these issues by enhancing the robot navigation through a collaborative control method that assists the user to negotiate obstacles. The method has been implemented in a commercial telepresence robot and a user study has been conducted in order to test the suitability of our approach.

The development of mobile context-aware applications using sensors require the developers to understand several diverse issues: signal acquisition, network protocols, embedded systems, data filtering, etc. We designed and implemented a software framework in order to assist developers in prototyping. Our framework facilitates the use of sensors from wearable devices and supports the reusability of components following a modular approach. This paper describes the design of our approach and highlights the benefits of the framework for the development of mobile applications. To evaluate the framework, representative context-aware applications are described as a case study. The usability of the applications were tested with 26 participants and good results were obtained.

Digital public displays have a unique capability to enable situated shared experiences, especially when open to user-generated content from people in their vicinity. The challenge, however, is how to open public displays to user-generated content, while being able to efficiently support conformity with place and display owner expectations. Sharing the display with users potentiates many risks, which go far beyond the feared appropriation of the display for presenting offensive content. In this study, we conducted a systematic elicitation of the risks involved. Based on a qualitative analysis of moderation situations referred in the literature, we identify and describe 7 specific risks that display owners should manage to be able to support user-generated content.

We present the design, development and evaluation of a haptic enhanced tele-rehabilitation system for massage therapy of the back using the Vybe haptic gaming pad. The proposed haptic system includes features that allow (i) administering online therapy programs, (ii) self-adjustable and safety treatment of back massages using a virtual environment, the gesture sensor LEAP Motion controller and the Vybe haptic device, and (iii) save and replay messages according to the therapy program. A usability evaluation with 25 elders suggests that the haptic tele-rehabilitation system is perceived as relaxing, useful and usable, while providing a supervised, real-time and secure way to treat the patient and adjusting the therapy haptic feedback intensity.

Usability testing of mobile applications involving people with Down syndrome is an issue that has not been comprehensively investigated. There is no single proposal that takes on board all the issues that could potentially be taken into account to deal with the specific needs of people with Down syndrome. We propose a guide for a usability testing process involving participants with Down syndrome. This guide is called USATESTDOWN. It is based on a literature review and experience gained at a number of workshops where people with Down syndrome used mobile devices. This paper briefly describes USATESTDOWN and its application at a special employment centre called PRODIS with 10 participants.

Learnability is a fundamental usability factor. It is included in several well known and widely used software usability evaluation models such as: IBM’s Computer Usability Satisfaction Questionnaire (CUSQ), System Usability Measurement Inventory (SUMI), System Usability Scale (SUS), and others. However, all of them assess learnability only subjectively. Taking into account the fact of presence of differences in perceived duration in interaction between human and computer, development of a new approach targeting usability, as well as learnability, in objective measures could be of great interest. This paper describes our endeavour in pursuing this task. We present our approach to the problem, introduce an instrument for calculating objective learnability out of times of completion, and reveal results of a user study among 101 participants conducted to test adequacy of the method.

The irruption of computer-based technology in social interaction has negatively affected the way in which elderly people interact with their family members, because they are, in some cases, reluctant to adopt new digital media. Given that literature suggests that seniors spend a significant part of their day watching television, we argue that Smart TV applications can be an effective way to provide them access to ambient assisted living services. This paper reports the design and use of a Smart TV-based application that promotes social interaction between older adults and their family members through social media. The system runs on a LED screen, extended with smart functionalities provided by Google Chromecast. The social interaction features provided by the system include exchanging email messages and sharing photos that are automatically retrieved from the accounts of family members in social media feeds. The system was evaluated with a group of expert users as well as with a sample of end-users. Older adults participating in the study praised the new possibilities offered by the prototype application as a way to better engage with family-generated content, thus facilitating their social integration.

Assistive technologies have been used to improve the quality of life of people who have been diagnosed with health issues. In this case, we aim to use an assistive technology in the shape of an affective avatar to help people who have been diagnosed with different forms of Social Communications Disorders (SCD). The designed avatar presents a humanoid face that displays emotions with a subtlety akin to that of real life human emotions, with those emotions changing according to the interactions that the user chooses to perform on the avatar. We have used Blender for the design of the emotions, which are happiness, sadness, surprise, fear and anger, plus a neutral emotion, while Unity was used to dictate the behavior of the avatar when the interactions were performed, which could be positive (caress), negative (poke) or neutral (wait). The avatar has been evaluated by 48 people from different backgrounds and the results show the overall positive reception by the users, as well as the difference between neurotypical and non-neurotypical users in terms of emotion recognition and chosen interactions. A ground truth has been established in terms of prototypic empathic interactions by the users.

Analyzing activities (either static postures or movements) made by a user is a complex process that can be done through a wide range of approaches. One part of these existing approaches support doing the recognition focusing their analysis on specific body parts. In fact, in previous publications a method was introduced for activity recognition (Body-Angles Algorithm) capable of analysing only using a single sample of those activitites and allowing the selection for each activity which are the relevant joints. But being able to analyse the body of the user selecting only a subset of the same, has both advantages and disadvantages. Therefore throughout this article we will expose those disadvantages, the applied solution to mitigate them and the results of an evaluation destined to clear which body parts make it easier to obtain high accuracy rates in recognition. Through this work we aim to give the scientific community lessons learned about the usage of different body areas in the analysis of activity recognition.