Examinando por Autor "Paddy Junior, Asiimwe"

Mostrando 1 - 3 de 3
  • Miniatura
    Ítem
    Efficient remote pedestrian localization techniques for resource-constrained environments
    (Universidad de Deusto, 2025-07-24) Paddy Junior, Asiimwe; Díez Blanco, Luis Enrique; Eyobu, Odongo Steven
    This thesis explores resource-efficient pedestrian localization techniques tailored for outdoor remote resource-constrained environments, specifically focusing on enhancing remote monitoring for elderly people. With a global trend toward aging populations, particularly in underdeveloped regions, and a policy shift from institutional care to community-based aging, it is estimated that by 2050, two-thirds of the world’s population over 60 will reside in low- and middle-income resource-constrained countries. This trend is likely to place a higher burden on healthcare, family, and social services since many of these elderly people cannot live independently without assistance from a caregiver. Remote monitoring systems offer promising solutions to bridge the gap between elderly individuals’ needs and available healthcare services, but their adoption is still limited in environments that lack basic infrastructure like stable power and communication networks. Therefore, this study begins by clearly defining ”resource-constrained environments”and systematically reviews existing outdoor remote pedestrian localization systems, evaluating their suitability for these environments. Global Navigation Satellite System (GNSS) technology is highlighted as the most viable option for remote, accurate, long-range, infrastructure-free outdoor localization, but its high power consumption presents challenges when integrated into battery-powered, wearable IoT devices. This research proposes two methods for efficient GNSS activation specifically tailored to resource-constrained environments to make GNSS feasible. Both approaches aim to minimize unnecessary GNSS activation, optimizing power consumption while ensuring reliable user localization. The first method is a position-based GNSS activation approach using a Pedestrian Dead-Reckoning (PDR) system. By leveraging a predefined geofence (safe zone) around the user’s home, GNSS activation is minimized, turning on only when the user leaves the safe zone. Two implementations of the PDR system were evaluated: accelerationbased and pitch-based. The proposed PDR-based activation method was validated through extensive experimental evaluation, demonstrating significant improvements in power efficiency—up to 90% compared to acceleration-based methods both inside and outside the geofence without requiring costly infrastructure such as beacons. The second approach employs machine learning (ML) to drive GNSS activation based on user activity. By leveraging data from inertial sensors, the ML-based system differentiates between everyday ”at-home” activities and ”walking away” from home. The model activates GNSS only when it detects the specific user activity of ”walking away” from home. Four machine learning models were evaluated—Long Short- Term Memory (LSTM), XGBoost, Support Vector Machine (SVM), and Random Forest (RF)—with XGBoost being selected for implementation due to its strong balance between accuracy and computational efficiency. This method proved effective in significantly reducing power consumption, achieving over 40% savings compared to the acceleration based method. Experimental validation demonstrated that the proposed PDR-based and ML-driven GNSS activation methods are suitable for deployment in realworld, resource-constrained environments. By selectively activating GNSS, these methods effectively extended battery life, allowing wearable systems to support reliable remote monitoring for longer periods. The proposed solutions successfully addressed the goal of developing resource-efficient localization systems for aging populations, proving to be scalable and cost-effective while enhancing safety for elderly individuals and reducing the burden on caregivers. The targeted power optimizations make these systems feasible for current use and adaptable for future improvements, providing practical, energy-efficient remote monitoring solutions that help elderly individuals age safely in their communities with minimal reliance on resource-intensive infrastructure.
  • Miniatura
    Ítem
    ML-driven user activity-based GNSS activation for power optimization in resource-constrained environments
    (Institute of Electrical and Electronics Engineers Inc., 2025-08-11) Paddy Junior, Asiimwe; Díez Blanco, Luis Enrique; Bahillo, Alfonso; Eyobu, Odongo Steven
    The aging population represents an increasing burden on healthcare systems, which is shifting policies from institutionalization to aging in the community. Remote monitoring offers efficient solutions that bridge the gaps between healthcare and where elderly people really want to live every day. However, the adoption of such systems remains low, especially in resource-constrained environments like underdeveloped regions and rural areas, due to the lack of resources often taken for granted in system design. Location is one of the main types of information to monitor, as it provides information about behavior and physical activity. Global Navigation Satellite System (GNSS) is the de facto technology, and although its high-power consumption aligns poorly with battery-powered devices, it is still the best choice for accurate and reliable remote localization of pedestrians. Deciding when to turn on/off the GNSS receiver based on context is a key strategy for power optimization, the two main types of contexts being the user’s position and activity. However, existing methods in the literature are not suitable for resource-constrained environments because they require the installation of beacons, which entail additional cost and power consumption, or assume the availability of external signals that are not met in such environments, or are based on simple user activity detection. This work proposes a new GNSS activation method based on detecting the specific walking activity for changing locations. In resource-constrained rural environments, people typically spend most of their time outdoors near their houses, where it is not necessary to activate the GNSS so frequently to monitor them. Restricting the GNSS activation to the moments in which they are moving to a different location could be enough and would reduce the power consumption. Four machine learning (ML) classification models [long short-term memory (LSTM), extreme gradient boosting (XGBoost), support vector machine (SVM), and random forest (RF)] have been implemented and evaluated using a smartwatch’s inertial sensor data. The best model, XGBoost, was exported to a custom-designed embedded system and evaluated in real-world tests. It demonstrated over 40% power savings compared to conventional motion-based methods.
  • Miniatura
    Ítem
    Remote pedestrian localization systems for resource-constrained environments: a systematic review
    (Institute of Electrical and Electronics Engineers Inc., 2023-04-13) Paddy Junior, Asiimwe; Díez Blanco, Luis Enrique ; Bahillo, Alfonso ; Eyobu, Odongo Steven
    The steady increase in the number of elderly citizens represents a likelihood of an increased burden on the family, government, healthcare, and social services since many of these elderly people cannot live independently without assistance from a caregiver. As such, there is an increase in demand for services in terms of technologies to address the urgent needs of the aging population. Remote monitoring, which is based on non-invasive, non-intrusive, and wearable sensors, actuators, and communication and information technologies, offers efficient solutions that bridge the gaps between healthcare and where elderly people really want to live every day. The rate at which such platforms have been adopted is extremely low in low-developed countries and rural areas, one of the main reasons being the lack or scarcity of some resources that these systems take for granted. In other words, these systems are designed for developed countries but are very much needed in resource-constrained environments as well. This study provides an in-depth, state-of-the-art systematic review of the current outdoor remote pedestrian localization systems to identify their suitability for resource-constrained environments. After checking 35 survey papers from the last ten years to the best of our knowledge, this is the first survey that investigates the suitability of existing pedestrian localization systems for a resource-constrained environment. This study is based on PRISMA guidelines to provide a replicable work and report the studies' main findings. A total of 37 works published between 2012, and January 2023 have been identified, analyzed, and key information that described the devices and tools used, communication technologies, position estimate technologies, methods, techniques and algorithms, and resource optimization strategies currently used by the localization systems was extracted to help us answer our question. The results indicate they are not fit for a resource-constrained environment as most assume the availability of infrastructures such as Wi-Fi, Internet, cellular networks, and digital literacy, among others, for their systems to operate properly, which are limited or not available in the resource-constrained environment described in this review.