Team project-based learning (TPBL) combines two learning techniques: 1) project-based learning (PBL) and 2) teamwork. This combination leverages the learning outcomes of both methods and places students in a real work situation where they must develop and solve a real project while working as a team. TPBL has been used in two advanced database subjects in Jaume I University (UJI)’s Computer Science degree program. This learning method was used for four years (academic years from 2018/2019 to 2021/2022) with positive outcomes. This study presents the project development, which includes teamwork formation, activities, timetable, and exercised learning competencies (both soft and specific). Further, the project’s results were evaluated from three different perspectives: 1) teamwork evaluation by teammates; 2) students’ opinions on the subject and project; and 3) subject final grades.

MQTT is being widely used for data delivery in IoT applications but its architecture does not properly handle mobility when disconnection periods tend to be large. In this paper we describe an experimental evaluation, made in a real environment, of a solution that guarantees that there is no information loss when variable length hand-offs appear due to the movement of a node. Our proposal modifies the classical publish/subscribe scheme by introducing an intermediate buffer that takes care of message transfer. Finally, we study the impact related to the connectivity of mobile devices of the use of the standard LinuxNetwork Manager. We propose a cross-layer solution that improves the device connectivity in conjunction with the data layer management. We show that our solution improves the data delivery guaranteeing that no information is lost.

Riding a bicycle is a great manner to contribute to the preservation of our ecosystem. Cycling helps to reduce air pollution and traffic congestion, and so, it is one of the simplest ways to lower the environmental footprint of people. However, the cohabitation of cars and vulnerable road users, such as bikes, scooters, or pedestrians, is prone to cause accidents with serious consequences. In this context, technological solutions are sought that enable the generation of alerts to prevent these accidents, thereby promoting a safer city for these road users, and a cleaner environment. Alert systems based on smartphones can alleviate these situations since nearly all people carry such a device while traveling. In this work, we test the suitability of a smartphone based alert system, determining the most adequate communications architecture. Two protocols have been designed to send position and alert messages to/from a centralized server over 4G cellular networks. One of the protocols is implemented using a REST architecture on top of the HTTP protocol, and the other one is implemented over the UDP protocol. We show that the proposed alarm system is feasible regarding communication response time, and we conclude that the application should be implemented over the UDP protocol, as response times are about three times better than for the REST implementation. We tested the applications in real deployments, finding that drivers are warned of the presence of bicycles when closer than 150 m, having enough time to pay attention to the situation and drive more carefully to avoid a collision.

There is an important growth of unmanned aerial vehicles (UAVs) performing planned missions in urban environments, which poses significant challenges to the research community. The possibility of collisions represents a critical challenge. UAVs can suffer collisions due to different causes external or internal to their flight plans. In this context, dynamic geo‐fencing is a useful approach, whereby each UAV is able to provide a prediction of its future positions within a limited time. These predictions could be used to detect conflicts, allowing to dynamically modify the flight plans so as to avoid imminent collisions. In this work, a conflict detection algorithm/method is proposed, implemented and tested on a central server performing real‐time conflict analysis for a large number of UAVs flying in the aerial space of a city (U‐Space). The architecture assumes that UAVs send their future locations to a traffic controller. This controller compares the predicted positions of nearby vehicles to detect possible conflicts. The results of this work demonstrate the feasibility of the proposed conflict detection algorithm and its interest to improve the security and efficiency in U‐Space environments. The server is able to track thousands of UAVs in real time with a conflict anticipation around 11 s.

Building telecommunication network infrastructures is a key issue both, in developing regions and isolated areas in order to facilitate people not only access to information and new technologies, but also to give them the opportunity of self-organize. Nevertheless, despite the huge importance of the network infrastructure for the development of people, building a neutral and open communication system is not a priority for governments, so community networks are a solid and real alternative. In this document we present guifi.net, a bottom-up community network, born in Spain, having over 30,000 active nodes, totaling more than 55,000 km in network links, and showing a sustained growing rate. guifi.net is an opportunity to build a distributed telecommunications infrastructure governed by the participants in an indiscriminate way. But a network such as guifi.net is not only of interest in developing areas. In many countries, internet connections, despite being a basic service, are usually monopolized by a small number of private companies. An open and free network as guifi.net offers users the opportunity of being participants in the decisions regarding infrastructure and services.

The ageing of the population needs the automation of patient monitoring. The objective of this is twofold: to improve care and reduce costs. Frailty, as a state of increased vulnerability resulting from several diseases, can be seen as a pandemic for older people. One of the most common detection tests is gait speed. This article compares the gait speed measured outdoors using smartphones with that measured using manual tests conducted in medical centres. In the experiments, the walking speed was measured over a straight path of 80 m. Additionally, the speed was measured over 2.4 m in the middle of the path, given that this is the minimum distance used in medical frailty tests. To eliminate external factors, the participants were healthy individuals, the weather was good, and the path was flat and free of obstacles. The results obtained are promising. The measurements taken with common smartphones over a straight path of 80 m are within the same order of error as those observed in the manual tests conducted by practitioners.