Urban Computing Lab

The navigation and orientation of visually impaired people in urban environments are research topics that emerge within the broader area of inclusive mobility, following the recommendation of the Sustainable Development Goals for 2030. Inclusive mobility can include people with permanent or temporary mobility disabilities such as visually impaired people, people with autism spectrum disorder, deaf people, people in wheelchair, pregnant ladies, elder or people with toddler. Each one of these segments has different specificities considering their mobility.<BR> <BR> Regarding visually impaired people, one of the main challenges they face in an urban environment is to obtain a point of reference when they lose track of their whereabouts. One of the approaches that can have promising results in this context are ontologies which allow the classification and relation of concepts of a given domain. This usage can allow to extrapolate the user's precise position through an image captured by the device's camera, based on a dataset of images that represent significant points of interest in a city. This approach can also take advantage of sensory information in the environment – such as smell and hearing – to help in the positioning inference.<BR> <BR> When considering all segments of mobility disabilities, obtaining personalized routes is also still a challenge, mainly in high density areas. Obtaining optimized routes to reach a desired destination is already a common practice that citizens are used to and that greatly facilitates their travel in an urban environment. However, some people have specific mobility needs, which implies that the recommended routes need to take into account different information such as the slope of the streets, the existence of crosswalks, loud noises, the existence of emergency exits, among other aspects. It is therefore necessary to have algorithms adapted to these situations, which take into account multiple criteria and which can respond positively in terms of performance in real-time use, when inserted into mobile applications. The scalability of a solution of this nature must also be considered, as the classification of street segments in large cities becomes a challenge. The contribution of citizens, from a crowdsourcing perspective, represents an additional challenge to the scalability of the solution and the optimization of the algorithms.<BR> <BR> Finally, Mobility-as-a-Service (MaaS) solutions’ importance is steadily growing given the focus cities have on ensuring and promoting sustainability. One of the main goals of MaaS solutions is the recommendation of multi-modal routes to reduce the private vehicle by promoting shared transport options. However, as with other mobility solutions, MaaS solutions must also consider universality and accessibility. If we think about the segment of people with permanent or temporary mobility difficulties, a set of challenges arise that must be addressed by MaaS solutions, translating into an iMaaS (Inclusive Mobility-as-a-Service). Some of these segments include people on the autism spectrum, people who are blind or visually impaired, people who are deaf, in wheelchairs, pregnant women or adults with infants. Each one has specificities that must be treated in a unique way in any MaaS solution, promoting greater autonomy in these segments.<BR> <BR> There are numerous challenges that are currently posed to inclusive mobility so that it can allow universal and accessible access to all citizens without exception, as recommended in the Sustainable Development Goals for 2030, which makes this a promising area for research contributions.

The rapid and facilitated access to information promoted by technological evolvement, supports citizens to be better informed and have a more active participation in governance issues within the city they live, develop their professional activity or participate in cultural and social events. On the other hand, municipalities demonstrate openness to the external community concerns, encouraging their participation in decisions related to the local government. An important and global reference is the Participatory Budgeting (Orçamentos Participativos) aimed to invite citizens to present solutions to the real needs and community problems, for later financing. Some developed studies highlight the added value of the citizen engagement in local governance, namely in improving the provision of public services, in managing public finances, in contributing transparency or increasing accountability and social inclusion. In fact, and despite of being promising for the future of cities, this is a research area that still remain in a very embryonic stage of development. There a small number of examples, mainly focused in some specific processes and tools. A true transformation requires a much broader change, introducing the paradigm concept “city as a service”. This change implies identifying fundamental services that can be easily combined with each other to provide more complex functionalities, but designed with an integrated logic and strongly focused on the citizen’s needs.<BR> <BR> In this context, this research proposal aims to identify and specify this type of services in order to generate new knowledge to effectively understand how to approach this new perspective, particularly in connecting the city with the citizen’s engagement.<BR> In a first stage the project involves case studies of successful services, such as Participatory Budgeting, applications to address the problems or mobile applications to perform parking payments. In a second phase, the research project should comprise the characterization of the urban services ecosystem in order to enable the identification of the involved players, as well as their participation’ scope, and define a system model to support an integrated approach without having a closed system. The formalization of this model should generate a new and systematic approach, to support the new service development, reducing costs and promoting the user’s participation through the use of well-known regular tools and processes instead of the common multiple ad-hoc mechanisms. <BR> The main goal is to encourage the active digital participation of citizens in local governance and enable them to recognize their efforts and making them feel involved. In practice, it translates into a mindset change: “citizens as active agents, rather than passive recipients”.

1. Cortés-Cediel ME, Cantador I, Bolívar MPR. Analyzing Citizen Participation and Engagement in European Smart Cities. Social Science Computer Review. 2021;39(4):592-626. doi:10.1177/0894439319877478. Available in: https://journals.sagepub.com/doi/pdf/10.1177/0894439319877478.<BR> <BR> 2. Elelman Richard, L. Feldman David. The future of citizen engagement in cities—The council of citizen engagement in sustainable urban strategies (ConCensus), Futures. Volume 101, 2018, Pages 80-91. ISSN 0016-3287, https://doi.org/10.1016/j.futures.2018.06.012. Available in: https://www.sciencedirect.com/science/article/pii/S0016328717301763.<BR> <BR> 3. Portney, K. Civic Engagement and Sustainable Cities in the United States. Wiley Online Library. Volume 65, Issue 5, Pages 579-591, https://doi.org/10.1111/j.1540-6210.2005.00485.x

It is commonplace to say that we can get the information we want at the distance of a click. Still, the availability of large amounts of information does not mean we can always get the right information, at the right time. Information relevance can be strongly situated, contextual and spontaneous, evolving dynamically as we intermix our daily life tasks. To be relevant, information should be strongly aligned with these situated needs and readily available. For example, imagine being at a bus stop and trying to find the time until the arrival of the next bus, or, being at a major urban festival and trying to know the next shows taking place at a particular stage. Both examples refer to information that is certainly available on the Web, and yet there is currently no generic and scalable approach to explicitly make such digital content readily available in those contexts where it is most needed. Existing alternatives, e.g. location-based applications, public displays or augmented reality, are all based on custom solutions and closed system designs, which severely undermine their widespread adoption.<BR> This project explores a novel approach to locative media, based on web technologies and linked-data, but especially on the same design principles behind the success of the web. It also takes a unified perspective on locative media, embracing location-based applications, public displays or augmented reality as complementary channels in a global sprawling physical-virtual hyper-system. <BR> <BR> The project is highly innovative because no such system exists today and, if successful, it will represent a radical transformation in how we think about locative media. The respective research may address various topics around system design, interaction or the in-depth exploration of specific application domains.

<strong>Motivation</strong> <BR> Many different applications can benefit from indoor positioning solutions. For some specific cases, like controlling the movement of an industrial machine (e.g. a forklift), it is mandatory to have high accuracy positioning estimates to execute an action/movement safely.<BR> Ultra-Wide Band (UWB) is the most well-known technology deployed to implement high-accuracy solutions indoors. Still, it lacks scalability at a reasonable price and demands a dedicated infrastructure that is not easy to deploy in some places. LiDAR - Light Detection And Ranging and camera-based solutions allow to tracking the movement of vehicles and other elements in the space to ensure security in motion. Although these types of sensors have also been explored to determine absolute position, they have severe practical limitations.<BR> <BR> <strong>Research Topic</strong> <BR> This proposal encompasses the development of a low-cost, highly accurate solution for indoor positioning based on infrared emitters/receivers. The goal is to explore implementing a solution that will use a (set of) narrow infrared beam emitter(s) and a set of receivers with several infrared detectors. The Angle-of-Arrival (AoA) estimation from several receivers will be combined to estimate the emitter's position.<BR> This proposal is part of a research program submitted for funding, encompassing researchers from three different institutions. A research grant may become available within the project funding.

Many different applications can benefit from indoor positioning solutions. For some specific cases, like controlling the movement of an industrial machine (e.g., forklift), it is mandatory to have high accuracy positioning estimates to execute an action/movement safely.<BR> Ultra-Wide Band is the most well-known technology deployed to implement high accuracy solutions. Still, it lacks scalability at a reasonable price and demands a specific infrastructure that is not easy to build in some places. LiDAR - Light Detection And Ranging and camera-based solutions support the movement of vehicles to ensure security in motion. Although they can also determine absolute location, they have severe limitations.<BR> This proposal encompasses the development of a low-cost, highly accurate solution for indoor positioning based on infrared emitters/receivers. The goal is to explore implementing a solution that will use a (set of) narrow infrared beam emitter(s) and a set of receivers with several infrared detectors. The angle-of-arrival estimation from several receivers will be combined to estimate the emitter's position.<BR> This proposal is part of a research program submitted for funding, encompassing researchers from three different institutions. A research grant may become available within the project funding.

<strong>Motivation</strong><BR> Data is one of the most valuable assets in the digital era. Data Science and Machine Learning are currently two of the most popular areas with high demand for large volumes of data. In many areas, obtaining real-world data, is time-consuming and costly and may involve privacy issues. Synthetic data, is artificially generated and is emerging as a valuable resource to replace or complement real-world data. Synthetic data can also be used to test a specific component or condition, but also complex systems. As an example, in aeronautics, complex and very specific simulators are used to create virtual scenarios and events to test systems in different use cases and conditions. In the Indoor Positioning and Navigation (IPN) area, synthetic data is also of utmost importance, for example to train algorithms or for validation of the system’s performance. <BR> <BR> The Dioptra project focus in helping the academic community and the industry to generate high quality synthetic data to test and validate their IPN solutions. The tool is based on a modular architecture that allows to extend their capabilities and integrate new models and positioning technologies. The application provides a modern Graphical User Interface and is powered by a simulation engine that integrates all modules (sensors, actors, infrastructures) in a virtual environment.<BR> <BR> <strong>Doctoral Research Topic</strong><BR> Visible Light (VL) Based Positioning and Tracking is emerging as a promising solution for indoor environments. VL Positioning uses the free-space Optical Channel (OC) to transmit visible light signals. Light Emitting Diodes (LED) can be used to transmit these signals, therefore LEDs can be used to provide the positioning infrastructure while also provide illumination to the building. VL positioning uses one or more light sources (transmitters), an image/light sensor (receiver) and a Line-Of-Sight (LOS) communication channel. Different techniques can be applied to estimate the position of a device, including trilateration, triangulation or Time-Difference-Of-Arrival (TDOA). <BR> <BR> The research work of the candidate will be focused in the simulation of light transmitters (LEDs) and a receiver and the modulation of the optical channel/ light propagation, considering the 3D indoor environment. Different algorithms and methods can be explored to simulate/modulate the VL system components, such as the Ray Tracing technique, frequently used for game graphics rendering, realistically simulating the lightning of a scene, including the different light propagation effects (e.g. reflections , refractions, etc).<BR> <BR> The successful candidate will contribute to take Dioptra to the next level, researching and developing models and algorithms for state of the art technologies, such as: Optical Channel Modelling/ Visible Light Propagation; new Virtual Sensors (image/light receivers); a new 3D Simulation Engine.

<strong>Motivation</strong><BR> Multiple solutions to estimate and track the position of a person or device inside an indoor space have been proposed in recent years. Since GNSS (Global Navigation Satellite Systems) is not viable in these types of environments, other technologies, such as Wi-Fi and Bluetooth, started to be explored. With simple technical solutions based on Wi-Fi and Bluetooth it is easy to estimate a position with relatively low accuracy, such as with room-level accuracy, and that is enough for many applications. On the other hand, the indoor characteristics and the complex signal propagation effects (e.g. multi-path, fading, shadowing, etc) severely limit the positioning performance that is possible to achieve. However, with emerging areas such as the Internet-of-Things (IoT) and the Industry 4.0, the performance requirements increased both in terms of accuracy and reliability. Technologies such as Ultra-Wide-Band (UWB) started to become popular since they can achieve higher accuracy. The higher frequency and lower range of UWB signals result in less degradation of the signal, but this type of solution requires more LOS and the deployment of dense and dedicated infrastructure, resulting in higher cost which limits its widespread adoption. This is the reason why less accurate but more ubiquitous signals, such as Wi-Fi, that are already available in the indoor spaces, can still be a good option if the accuracy limitations can be overcome.<BR> <BR> <strong>Doctoral Research Topic </strong><BR> Wi-Fi signals can be explored in different ways to estimate a position. The most well-known approaches, such as the fingerprinting techniques, explore the Received Signal Strength Information (RSSI) (or signal power). However, this type of signal characteristic is severely affected by the propagation effects and the lack of Line-Of-Sight. With the recent advances in Wi-Fi Network Interface Cards is now easier to obtain the Channel State Information (CSI). The CSI provides the channel variations while the signal propagates from the source to the receiver (multipath effect, fading, shadowing, and delay distortion). This information is usually more stable than the RSSI and can be explored to achieve higher positioning performance.<BR> <BR> This PhD work will focus on researching and developing a new positioning solution that explores the Wi-Fi Channel State Information (CSI). The work will include a practical component of R&D, to develop a dedicated positioning tag and the application of different positioning algorithms and methods. <BR> <BR> The successful candidate will contribute to a new high-accuracy positioning system targeting industry applications with demanding accuracy requirements, for example, to support autonomous robots and machine navigation in Industry 4.0.<BR> <BR> <strong>A scholarship might be available to support the PhD student.</strong>