In recent years, companies have increasingly invested in Industry 4.0 research to automate repetitive tasks or activities that may be harmful to humans. For instance, large-scale warehouses, such as those operated by Amazon, rely heavily on mobile robots to streamline logistics operations and ensure worker safety. At the same time, robot manufacturers are releasing more reliable platforms with advanced capabilities, making them suitable for complex industrial tasks. Despite this growing interest and technological progress, significant challenges remain. A key and non-trivial issue lies in the integration of heterogeneous robot fleets: each vendor typically employs proprietary technologies and interfaces, which hinders interoperability and limits the potential of multi-brand deployments.In this work, we introduce HarmoniKt, an extensible middle-ware that addresses this challenge by introducing an abstraction layer and providing a unified REST API for the control and management of heterogeneous robots. Our solution has been validated in a physical industrial-like environment using a mixed fleet of Boston Dynamics Spot and Mobile Industrial Robots (MiR). Furthermore, we present a comparative analysis showing that the access latency introduced by our middleware is not significantly higher than that of direct robot access, demonstrating the feasibility of unified robot management without compromising performance.

The increasing deployment of smart devices in residential environments opens new opportunities for intelligent energy management. However, existing platforms often fall short in providing intuitive interfaces, zone-level control, and advanced predictive analytics accessible to non-expert users. This paper presents the design of a modular smart home energy management system that integrates real-time monitoring, consumption forecasting, and intelligent assistance via Large Language Models (LLMs). The system features an interactive floor plan interface, multi-user support, threshold-based alerting, and detailed historical analytics. Additionally, it introduces LLM-powered agents that guide users in configuring smart devices and adopting more efficient consumption behaviors. This architecture emphasizes accessibility, adaptability, and extensibility, aiming to empower users with actionable insights and seamless device management. The proposed solution addresses current gaps in existing platforms and lays the groundwork for intelligent, personalized, and proactive home energy systems.

Road surface quality is a key determinant of mobility safety, comfort, and sustainability in modern cities. Traditional pavement monitoring relies on costly and sparse inspection campaigns, which limit scalability. Recent advances in Edge Artificial Intelligence (Edge AI) and the widespread adoption of smart-phones open new opportunities for low-cost, distributed, and near real-time assessment of road infrastructure. In this work, we investigate the feasibility of classifying road surface conditions using mobile sensors available on consumer devices, including accelerometers, gyroscopes, and GPS modules. We collected data from bicycles and cars using a custom web application, applied a feature extraction and preprocessing pipeline, and evaluated multiple machine learning models. Our results demonstrate that accurate classification of surface types is achievable, with XGBoost and Random Forest models outperforming alternatives, particularly in car-based deployments. These findings highlight the potential of Edge AI to enable scalable, comfort-aware mobility services and support municipalities in proactive infrastructure management.

The adoption of Digital Twin (DT) technologies in public transport systems, particularly bus networks, is gaining momentum as cities seek smarter, more responsive, and efficient mobility solutions. Enabled by advances in IoT, AI, and Big Data Analytics, DTs offer real-time monitoring, simulation, and optimization of transit operations. However, despite their potential, the application of DTs in bus-based public transport remains relatively underexplored and fragmented across the literature. This study presents a Systematic Literature Review (SLR) aimed at synthesizing current research on DT technologies in this domain. Specifically, it investigates architectural models, technological frameworks, and platform designs; examines how AI and machine learning models are integrated to support operational tasks; and analyzes the role of Human-Computer Interaction (HCI) in the design and usability of such systems. By identifying key trends, challenges, and research gaps, this work provides a structured overview of the current landscape. Furthermore, it outlines directions for future research in DT-enabled public transportation systems.

Crowd management is the problem of dealing with physical crowds of humans, e.g., for safety reasons or to improve services through crowd-awareness. Information and communication technology can be of great help in supporting the monitoring and management of crowds. Following the “digital twin concept”, a group or crowd of people – as a dynamic physical entity – is amenable to be represented and managed through digital models. Accordingly, in this work we present and develop a general notion of a Crowd Digital Twin (CDT) , intended as the digital twin associated to physical gatherings of humans. We motivate it as a fundamental component of crowd management systems, and investigate the functional and non-functional requirements that such a technical solution should satisfy. Then, we consider what the fundamental components of CDT systems are, and accordingly delineate a reference architecture aimed at flexibly supporting the corresponding physical-digital thread. We also discuss methods and tools that might be useful for supporting implementations of such a CDT concept, and delineate a set of challenges that may serve as a roadmap for future research on the topic.

Abstract As data visualizations become increasingly central to communication, analysis, and decision-making across a variety of fields, the need to ensure their accessibility becomes critical. Visual representations, such as charts, graphs, and infographics, are effective tools for conveying complex information; however, they often present significant barriers for individuals with various types of impairments. This paper explores the potential of Machine Learning (ML) to enhance the accessibility of data visualizations. Specifically, we present a Systematic Literature Review (SLR) that investigates how ML techniques have been applied to make visual data more inclusive. The review considers both visualization-related aspects (such as visualization types and source, target user, and evaluation) and ML-related factors (including data formats, preprocessing, and model types and evaluation). Our findings reveal that only a limited number of studies directly address the use of ML for improving visualization accessibility, and there is a lack of standardized solutions or frameworks in this area. Our contribution focused on the identification of a conceptual framework based on nine key open challenges that highlight the lack of consideration for infographics, particular types of charts, printed chart, different kind of impairments, involvement of users, accurate interpretation of complex visual data, real-time support, standard benchmarking, the potential bias and, hence, the need for continued research and development at the intersection of data visualization and machine learning, with a strong focus on accessibility and inclusivity.

Crowd management is the problem of dealing with physical crowds of humans, e.g., for safety reasons or to improve services through crowd-awareness. In previous work, the notion of a Crowd Digital Twin (CDT), namely the digital twin associated to physical gatherings of humans, was motivated and proposed as a fundamental component of crowd management systems. In this work, we better characterise the CDT notion, by investigating the functional and non-functional requirements that such a technical solution should satisfy. Then, we consider what are the fundamental components of CDT systems, and accordingly delineate a general architecture aimed at flexibly supporting the corresponding physical-digital thread.

The development of a crowd digital twin offers significant potential for enhancing public safety, urban planning, and event management. A key challenge in creating such a digital twin lies in the efficient and accurate acquisition of crowd-related data, particularly through object detection models deployed on resource-constrained devices. Through a series of experiments, we compare TinyML and Edge approaches in terms of detection accuracy, inferencing time, and resource utilization. Our findings highlight the trade-offs inherent in selecting detection models for crowd digital twin applications, underscoring the importance of aligning model choice with specific deployment needs.

Urban environments face significant challenges due to deteriorating road pavements, affecting all transportation modes' safety and comfort. Traditional methods of road assessment are costly and infrequent, but advancements in mobile technology and crowdsensing offer real-time, large-scale data collection solutions. This paper introduces “ShakeSensing”, a mobile application that uses Android device sensors to measure vibrations experienced by cyclists and scooter users, providing insights into road quality. The collected data, analyzed to assess road conditions, can benefit both users and municipal authorities, promoting safer and more comfortable travel.

This paper explores the evolving concept of digital sustainability, emphasizing its growing importance as digital technologies become essential to addressing global sustainability challenges. Through a thematic literature review focused on computer science, we analyze papers from Scopus-indexed journals to identify key themes in the discourse on digital sustainability. Additionally, we map the identified topics to the traditional three-pillar model of sustainable development, noting areas of alignment and divergence. The study offers insights into how digital sustainability is framed within the computer science literature and its implications for future research.

Accurate demand forecasting is essential for optimizing public transportation systems, reducing congestion, and improving service efficiency. Traditional methods often struggle to capture the dynamic nature of urban mobility, leading to increasing reliance on machine learning and deep learning techniques. This paper explores various machine learning models for bus transportation demand forecasting, leveraging real-world data to evaluate their effectiveness. Our findings contribute to the development of intelligent public transport systems, supporting data-driven decision-making for adaptive and sustainable urban mobility solutions.

Human-robot interaction (HRI) is an evolving field with a growing emphasis on enabling robots to understand and perform tasks based on natural language commands. Recently, Large Language Models (LLMs) have emerged as a promising tool for such tasks, offering the potential to enable zero-shot learning and flexible interaction without task-specific training. In this paper, we explore the use of LLMs for zero-shot navigation and exploration tasks in robotic systems, specifically evaluating their performance with the PR2 Clearpath and Khepera IV robots in a simulated environment. Our findings demonstrate promising results, particularly in the LLM’s ability to exhibit exploratory behavior and iterative reasoning when faced with ambiguous or incomplete visual input. These capabilities suggest a strong potential for LLMs in human-robot interaction. However, challenges were also identified, such as difficulties with target recognition, object misidentification, hallucination of information, and issues with movement execution, highlighting the need for improvements in these areas for real-world applications.

The widespread diffusion of Large Language Models (LLMs) has ushered in a transformative era across numerous research domains, including web accessibility. In fact, they can potentially offer automated solutions for generating accessible content, performing accessibility testing, and enhancing the overall user experience for individuals with disabilities. In this paper, we investigate how LLMs can be successfully employed to evaluate and correct web accessibility. Then, we delve into the positive implications and the current challenges derived from the interaction between developers and LLMs in this specific context. Finally, we present some future directions that could be explored to ensure that web content remains accessible to all.

Coastal lagoons serve as highly productive ecosystems, delivering essential ecosystem services contributing to human welfare and well-being. Unfortunately, these intricate systems are particularly vulnerable to climatic and anthropogenic pressures, such as intensive agriculture and extensive urbanization. This paper presents a Digital Twin prototype of the Mar Menor lagoon in Spain, that utilizes environmental data collected from sensors and predicted data generated through machine learning algorithms. It covers the prototype's architecture, the corresponding API layer designed for interoperability with other systems, and its data visualization dashboard aimed at enhancing decision-making processes.

Driving involves numerous factors demanding a driver's attention. To enhance road safety, there has been a significant increase in the development and implementation of vehicle sensors. These sensors, in conjunction with mobile applications, can assess a driver's emotional state and focus, providing feedback to enhance their attention. This paper explores the monitoring of driver behavior, emphasizing the effects of distractions and emotions on driving performance. Stemming from the European initiative, NextPerception, this research focuses on advancing perception sensors and refining distributed intelligence models in various domains, including automotive. The initiative aims to develop a range of sensors, from obstacle detection tools to those monitoring a driver's eye movements and vital signs. A key goal is to define a “fitness-to-drive” metric, representing the driver's attentiveness level. The project also seeks to use gamification to emphasize the importance of this metric, increasing drivers' awareness of their driving skills. The ultimate aim is to create a system that determines fitness-to-drive based on distractions and emotions, integrating this into a prototype simulating sensor data, and introducing a web-based application to display this data for a community of drivers.

Multimodal interfaces go beyond the traditional interaction through keyboard and mouse by incorporating multiple modes of interaction, such as touch, voice, gesture, and even gaze, to create more intuitive and immersive user experiences. This paper investigates how TinyML can be employed for multimodal interfaces in the context of games. An endless game in which the character has to avoid obstacles and fight monsters to advance has been developed. An Arduino Nano 33 BLE Sense is then used as the input device for the game by recognizing the hand gestures of the players.

The term Digital Twin (DT) refers to an emerging concept and technology extending the value of physical assets through the services supported by a bidirectional connection with their digital counterparts. Since its very introduction, the DT concept has been applied to empower several kinds of physical assets across multiple domains, including Industry 4.0, infrastructures, autonomous vehicles, e-health, to name a few. The rise of socio-technical systems and cyber-physical-human systems, which feature humans as key system components, has led to the extension of the DT concept to humans themselves, bringing in the notion of "human digital twin", supporting human-system integration. However, in several contexts, humans are not just isolated entities, but rather form groups or crowds. The peculiar dynamics of such groups have to be carefully considered in various scenarios like emergency management or logistics scenarios. Motivated also by recent trends in social Internet of Things and collective computing, in this paper, we propose the notion of a "crowd digital twin", meant to support services considering the crowd and for the crowd itself. Supporting a bidirectional connection with a physical crowd, this notion is shown to present peculiar requirements, opportunities, and challenges, including low-latency synchronisation, environment-driven prediction, actuation, and validation of emergent behaviour.

One of the main problems of web recommender systems is exposure bias, due to the fact that the web system itself is partly generating its own future, as users can only click on items shown to them. This bias not only creates popularity bias for products but also is one of the main challenges for recommender systems that deal with a very dynamic environment, where new items and users appear frequently and also user preferences change (or the market changes as happened with the coronavirus pandemic). The main paradigm to deal with these changes is to explore and exploit, avoiding the filter bubble effect. However, too much exploration also reduces short-term revenue and hence is usually traffic bounded. In this work, we present a counterfactual analysis that shows that web recommender systems could improve their long-term revenue if significantly more exploration is performed. This is good for the web recommender system but also for everyone as it creates more fair and healthy digital markets. This also improves the web user experience so is a double win-win for the e-commerce platform, the sellers, the users, and ultimately society.

Multistep tasks refer to activities of daily living that require a sequence of actions to be performed (e.g., preparing a sandwich). Engaging persons with intellectual disabilities in such tasks is an important educational goal, but these persons may find it difficult to independently execute the correct sequence of steps necessary to perform the intended activity. In this proof-of-concept study, we evaluated the usability of a web application running on a mobile device to enable three young adults with intellectual and motor disabilities to independently complete multistep tasks. The results of this study provide preliminary evidence about the overall usability of the web application for persons with disability and their caregivers. Future studies may build on current evidence to prove the overall effectiveness of the proposed system.

Gamification has been widely and often successfully exploited in several educational environments at different levels. In this study, we detail the use of Gamification mechanisms within a university course about web technologies, with the aim of increasing the students’ involvement in the study of the course subject. We designed different Gamification strategies based on group or individual activities. The study has been carried out over five academic years. Based on our findings, we can conclude that, under the defined conditions, the employed Gamification mechanisms had a positive impact on the students learning experience.

Passwords are the base of almost all authentication systems. Hence, it is essential that users choose strong passwords, to prevent attackers from guessing them by using the most common passwords. In this paper, we present a serious game developed with the aim of increasing the awareness of players on the (in)security of passwords. It has been developed using standard web technologies and employing the MEAN stack. To gain points, players have to guess which of the two proposed passwords is the most used one. Our prototype has three different game modalities both single and multi-player. A preliminary evaluation session has been conducted to evaluate the effectiveness of our prototype and its usability.

Social media are revolutionizing various areas of our society, including the cultural heritage one. The same goes for artificial intelligent strategies and machine learning algorithms, which are exploited in many contexts, and could play an interesting role in promoting and sharing cultural heritage and touristic activities and points of interests in urban environments. An interesting application of these techniques in the cultural and touristic domain could be represented by the automatic recognition and classification of monuments and points of interest. In this work, we present a study on monuments recognition, starting from pictures taken by users by means of their mobile devices, while visiting an urban environment. In particular, we employed images collected by the citizens and tourists with the aim of contributing to valorize monuments or places in a specific case study, that is the city of Cesena (Italy). In this paper, we compared several architectures of convolutional neural networks, that are able to discriminate among a set of outdoor monuments with an accuracy of over 90%.

As the world becomes increasingly interconnected, emerging and innovative sensing technologies are shaping the future of agriculture, with a special focus on sustainability-related issues. In this context, we envision the possibility to exploit Social Internet of Things for sensing of environmental conditions (solar radiation, humidity, air temperature, and soil moisture) and communications, deep learning for plant disease detection, and crowdsourcing for images collection and classification, engaging farmers and community garden owners and experts. Through, data fusion and deep learning, the designed system can exploit the collected data and predict when a plant would (or not) get a disease, with a specific degree of precision, with the final purpose to render agriculture more sustainable. We here present the architecture, the deep learning model, and the responsive Web app. Finally, some experimental evaluations and usability/engagement tests are reported and discussed, together with final remarks, limitations, and future work.

Indoor wayfinding systems are still presenting challenges and open issues in providing accurate directions while navigating complex and large buildings. When considering users who are experiencing any kind of vision loss, these issues can become really serious to tackle. To facilitate users (both visually impaired people and people without disabilities) in navigating an indoor environment, we designed and implemented a pervasive system taking advantage of the beacon technology and an accessible mobile application. In this paper, we present the system design issues and implementation challenges we addressed to provide user-friendly services in indoor contexts, exploiting the Proximity technique.

With the amazing success of PokemonGo, pervasive games have definitely caught the attention of the entire world. Their game experience is characterized by the fusion of real and virtual elements. In order to be able to extend the virtual world into the real one, such games have to access the mobile devices sensors like GPS and camera. For this reason, they are implemented as native or hybrid mobile applications. Contrary to this trend, in this paper we present Shoot Them All, a pervasive game implemented as a web application. In Shoot Them All, players have to hit the opponents, trying, at the same time, not to get hit. The battlefield is the real world, while the guns are virtual, as an extension of players' mobile devices. We were able to successfully developed such a game thanks to the availability of Javascript APIs that allow to access and manage device sensors. The success of this case study highlight how these technologies are mature enough, opening the door to web pervasive games.

In this paper, using Reddit, we investigated on the efficacy of using an AI-based mood analysis methodology to understand how the prescription of different prenatal diagnostic tests (invasive vs non-invasive), and their corresponding outcomes, may have an impact onto the mood of parents-to-be. We found an interesting causal relationship between tests and mood that deserves a deeper comprehension by means of healthcare professionals. We have also demonstrated that combining AI-based dialogue analysis methodologies with advanced statistics may bring to a smart model, suitable for transforming a large amount of empirical data into correct medical interpretations.

Machine (learning)-based techniques have made substantial advances recently, and there is a general suggestion that they will drive major changes in health care within a few years. Yet, we all suffer from the lack of precise comparative studies on the accuracy of machine-based interpretations of medical data. To fill this gap, in this paper we investigate on the efficacy of using an automated mood analysis methodology to understand how patients react to the prescription to take different kinds of prenatal diagnostic tests (invasive vs non-invasive) and to the corresponding outcomes, based on conversations developed on Reddit. Our study essentially provides answers to research questions concerning: i) the popularity of prenatal diagnosis, ii) the patients' sentiment about different prenatal tests, iii) the existence of a cause-effect relationship between prenatal testing and patients' mood, and iv) the type of dialogues held by patients and physicians on this topic. Nonetheless, a general result emerging from our research is that a machine-based decision loop for now still needs human involvement, at least to alleviate the tension between empirical data and their correct medical interpretation.

Mobility and ambient conditions are key factors in urban environments, affecting well-being and quality of life. In this context, sensors, smart mobility, networks, connectivity can play a significant and strategic role, being exploited with the aim of improving data and information available to public administration and to each citizen. In this way, they can be supported in having more sustainable and aware behaviours and in getting useful information and services, improving their daily activities. In this paper, we present a prototype of smart bike eco-system, designed with the aim of collecting, aggregating and sharing data about air pollution and about the urban environment, which can be exploited in a smart mobility context thanks to sensor and vehicular networks.

Moving is one of the most important issues as regards independence while conducting daily activities, in both indoor and outdoor environments. This is considered particularly true for people with disabilities and it should be guaranteed in order to enable and facilitate their integration and their independence. In this context, IoT could be a means to define and obtain smart and more accessible buildings. In this paper, we present UniSAS, a system which has been designed to improve the accessibility of buildings thanks to an architecture based on Raspberry Pi and users' mobile devices. The paper describes the system architecture and a prototype we have developed. Some users' scenarios are illustrated, so as to discuss the feasibility of our approach.

Moving across a University campus (outdoor, among the buildings, and indoor, among classrooms and offices) could represent a barrier for students with disabilities, affecting their independence while they conduct their daily activities. Providing support by means of smart phones thanks to location technologies can be a useful means of integration and inclusion, with the effect of facilitating also tourists, newbies, or freshmen. This is particularly true in those contexts where the universities are hosted in historical buildings in old towns, which is a typical situation in the European countries. This paper presents AlmaWhere, a system based on beacon technology, designed and developed with the aim of equipping students of the University of Bologna with an indoor navigation system, providing them support in finding classes, labs, and libraries, with a specific attention those users with disabilities. The paper describes the main design issues, the system architecture, and technologies we have analyzed in order to design a prototype and two scenarios that involve personas.

The wide diffusion of mobile devices, digital technologies and telecommunication providers and infrastructures greatly supports communication and social activities among people all over the world. This (r)evolution in communication could represent a great opportunity for those people who use assistive technologies due to some kinds of disability, but it could become a digital severe barrier at the same time. Assistive technologies supporting people with disabilities can be a strategic tool to enhance their inclusion, integration, and independence, in particular for persons with disabilities that involve more senses, such as deaf-blindness, which is the combination of blindness and deafness. Deaf-blind users can communicate by mainly exploiting the sense of touch. Focusing on this kind of communication, we have designed and developed GlovePi, a low-cost wearable device, based on a glove equipped with sensors, a raspberry-pi and mobile devices. In this paper, we present an improved version of the GlovePi system, which extends the form of communication, by supporting the many-to-many one, aiming to increase the inclusion of deaf-blind people in social life and daily activities.

In the context of wayfinding systems, this paper describes the design and development of an augmented smartphone-based navigation system where wearable output devices have been put to good use to improve the mobility experience of people with disabilities. In particular, we developed a system as a proof of concept, exploiting smart glasses and smart bracelets to covey information about the navigated path in an augmented way. We conducted a preliminary field study with a few users with mobility impairments to evaluate the feasibility of our system. Results are really encouraging considering the approach of using different ways to covey notifications related to the direction to follow. At the same time, the outcome reveals that the smart glasses we used in the prototype influenced negatively the perception of the experience.

This work presents a use case on multimodal urban paths in a smart mobility context. The proposed solution builds on the experience already matured and developed by the authors in different fields: crowdsourcing and sensing done by users to gather data related to urban barriers and facilities, computation of personalized paths for users with special needs, and integration of open data provided by bus companies to identify the actual accessibility features and estimate the real arrival time of vehicles at stops. In terms of functionality, the first “monolithic” prototype fulfilled the goal of composing the aforementioned pieces of information to support citizens with reduced mobility (users with disabilities and/or elderly people) in their urban movements. In this paper, we describe a service-oriented architecture that exploits the microservices orchestration paradigm to enable the creation of new services and to make the management of the various data sources easier and more effective. The manuscript demonstrates the effectiveness of the approach showing a successful use case of a service that take into account multimodal paths, by involving cyclists, bicycle lanes, and bike sharing services in a urban environments. Such a use case take into account the user's interface and interaction mechanisms, which are strongly affected by the context of use.