Centro de investigación en Mecatrónica y Sistemas Interactivos

Parent Organisation

Universidad Indoamérica

Country

Now showing 1 - 10 of 393

Some of the metrics are blocked by your
consent settings
Real-Time Industrial Risk Mitigation through an IIoT-Based Acoustic Alert System with Embedded Edge Processing and Remote Monitoring
(Springer Nature Switzerland, 2026)
Vargas-Bustamante, Miguel
;
Álvarez-Tello, Jorge
;
Cevallos-Sornoza, Angie
;
Zapata, Mireya
Environmental monitoring schemes have been transformed by the incorporation of IoT technologies in industrial environments, enabling a shift toward continuous monitoring models and immediate response to risk situations. However, many available solutions rely on highly complex and costly SCADA infrastructures, which limits their adoption in facilities with budgetary constraints. In this context, the present research posed as its central question whether it is possible to design and implement a low-cost IoT system, based on open-source technologies, capable of providing real-time environmental monitoring with immediate acoustic alerts and reliable remote notifications. To address this question, an experimental architecture was developed based on Arduino and ESP32 platforms, integrating temperature and gas sensors, a DFPlayer Mini module for customized acoustic feedback, and a wireless transmission scheme to a remote platform. The system was validated through controlled tests involving the exceedance of critical thresholds, evaluating detection accuracy, alert activation latency, and communication stability during continuous operation. The results showed a correct detection rate above 95%, an average alert activation time below 1.5 s, and remote transmission stability greater than 97%. In addition, an approximate 40% reduction in operator response time was observed compared to traditional manual schemes. Consequently, the study demonstrates that an open and scalable IoT architecture can achieve adequate performance levels for industrial safety, contributing a robust human–computer interaction approach and establishing a solid foundation for future extensions toward predictive and interoperable capabilities.
Some of the metrics are blocked by your
consent settings
Invasion Status, Distribution, and Environmental Preferences of Non-Native Ornamental Thunbergia Species (Acanthaceae) in Ecuador: An Emerging Threat to Tropical Montane Forests
(2026)
Ana Reyes-Hernández
;
Ileana Herrera
;
Anahí Vargas
;
Oleas Gallo, Nora Helena
;
Josue Alvarez
Species of the genus Thunbergia, native to Africa, Asia, and Australia, are widely cultivated as ornamental plants; however, their ability to escape cultivation and establish themselves in novel environments poses a growing threat to tropical forests. Here, we provide the first nationwide assessment of Thunbergia species occurring in Ecuador, integrating data from citizen science platforms, herbarium collections, and field surveys. We analyzed spatiotemporal patterns of occurrence, evaluated invasion status based on wild persistence and spread, and assessed environmental preferences using climatic niche analyses. Species distributions were further examined across land-cover types, conservation areas, and forest–non-forest interfaces. We confirmed the presence of five Thunbergia species in Ecuador, two of which also occur in the Galapagos Islands. All species were recorded both in cultivation and in the wild, indicating ornamental horticulture as the main introduction pathway for the genus, and occurrences were documented within 24 conservation areas. Thunbergia alata, T. fragrans, and T. grandiflora were categorized as invasive in Ecuador. Among them, T. fragrans exhibited broad environmental tolerance across bioregions. Wild occurrences were predominantly associated with human-modified landscapes but frequently occurred near forest edges, indicating ongoing encroachment into natural forests. These findings highlight the urgent need for preventive and targeted management strategies, particularly against T. alata, which represents an emerging threat to Andean forest ecosystems.
Some of the metrics are blocked by your
consent settings
Hybrid BLE–LoRa architectures for energy-efficient and resilient wireless sensor networks: Experimental validation and adaptive clustering strategies
(Springer Science and Business Media LLC, 2026-06-03)
Del-Valle-Soto, Carolina
;
Briseño, Ramon A.
;
Valdivia, Leonardo J.
;
Varela Aldas, José
;
Visconti, Paolo
Wireless Sensor Networks are increasingly deployed in mission-critical scenarios where resilience and energy efficiency are paramount. This paper presents a hybrid architecture combining Bluetooth Low Energy (BLE) and Long Range (LoRa) technologies to enhance both robustness and energy-aware performance in adversarial environments characterized by reactive jamming attacks. A comprehensive experimental testbed was developed, integrating BLE and LoRa nodes, a dual-protocol gateway, and a reactive jammer emulator. We introduce an adaptive clustering algorithm that performs energy-aware role assignment and jamming mitigation based on signal anomaly detection and multi-metric routing. To validate its effectiveness, we conducted an extensive time-series analysis on energy consumption, retransmission rates, and signal resilience under both mitigated and non-mitigated conditions across BLE-only, LoRa-only, and hybrid BLE-LoRa networks. The results show protocol-dependent performance trade-offs under the proposed mitigation algorithm. While LoRa-only and hybrid BLE–LoRa networks exhibit consistent reductions in energy consumption, retransmissions, and variability, the BLE-only configuration demonstrates improved resilience at the cost of a moderate increase in energy usage and retransmission activity due to clustering and control overhead. Notably, the BLE-LoRa architecture balances delivery assurance and energy efficiency while maintaining communication even under interference. Furthermore, we provide statistical modeling and hypothesis testing confirming the mitigation algorithm’s significant impact. These findings offer a critical empirical contribution to the design of resilient and energy-aware heterogeneous WSNs and demonstrate the viability of real-time adaptive mitigation strategies for emerging smart environments
Some of the metrics are blocked by your
consent settings
Immersive Digital Twin for Industrial Control Education
(Springer Nature Switzerland, 2026-08-15)
Andaluz, Víctor H.
;
Carvajal, Christian
;
Culqui-Tipan, Javier
This paper presents the development and implementation of an immersive Digital Twin for industrial control education, conceived as an experiential learning tool that integrates a real industrial control system, a three-dimensional virtual environment, and a real-time communication infrastructure. The proposed approach follows a layered architecture linking a smart tomato greenhouse as the physical process, a Siemens S7-1500 programmable logic controller, a human–machine interface for supervision and variable adjustment, a Unity-based virtual environment for immersive interaction, and an IoT platform for real-time data transmission and remote visualization. The Digital Twin synchronously replicates the physical process, enabling users to analyze the relationship between environmental variables, control decisions, and system response, as well as to identify normal and critical operating conditions. The system was evaluated through an experimental study with twenty undergraduate engineering students, considering usability and learning outcomes. The System Usability Scale yielded an average score of 82.5 (SD = 6.1), indicating high user acceptance. Learning performance, assessed through a test–retest strategy, increased from 55.3 (pre-test) to 76.8 (post-test), corresponding to a mean gain of 21.5 points after two weeks. Statistical analysis confirmed a significant improvement (p < 0.001) with a large effect size, demonstrating the positive impact of the immersive Digital Twin on understanding industrial control systems. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2027.
Some of the metrics are blocked by your
consent settings
Point Mass Planning and NMPC Tracking for Time-Optimal Agile Quadrotor Gate Navigation
(Institute of Electrical and Electronics Engineers (IEEE), 2026)
Guevara, Bryan S.
;
Brandao, Alexandre Santos
;
Varela Aldas, José
Time-optimal agile flight through a sequence of three-dimensional gates requires a consistent coupling between trajectory planning and tracking control. Point-mass planners offer fast numerical solutions for free-final-time optimisation, but the value of progressively enriching the reference fed to a nonlinear model predictive control (nmpc) tracker through a differential flatness bridge has not been clearly characterised on multi-gate circuits. This paper presents a modular three-stage pipeline: an offline time-optimal point-mass model (pmm) planner solved with CasADi and IPOPT; a differential flatness bridge that maps the flat outputs and their time derivatives to a full quadrotor reference comprising desired position, velocity, attitude quaternion, body angular rate, and collective thrust; and a body-rate nmpc tracker implemented on the acados SQP-RTI scheme with an SO(3) logarithmic-map external cost. Two reference configurations are compared in a software-in-the-loop setup that drives the MuJoCo physics engine through ROS 2 under a randomly seeded external-force disturbance. The baseline, nmpc-Att, uses only a yaw-only attitude reference derived from the velocity heading and carries no angular-rate or thrust feedforward. The proposed configuration, nmpc-Full, activates the complete flat reference, including the full attitude quaternion, body angular rate, and thrust feedforward. Results on a seven-gate figure-8 and an eight-gate vertical loop show that reference enrichment yields a categorical improvement in gate crossing and a systematic reduction of tracking error without increasing solver cost. The loop circuit further highlights a structural limitation of the common practice yaw-only reference: it cannot represent inverted attitudes, and only the flatness-based reference resolves this case
Some of the metrics are blocked by your
consent settings
Low-Cost IoT System with Containerized AI and Telegram Bot for Real-Time Air Quality Risk Communication and Preventive Behavior Change
(Springer Nature Switzerland, 2026)
Zapata, Mireya
;
Álvarez-Tello, Jorge
;
Rugel-Sanchez, Keyla
;
Vargas-Bustamante, Miguel
Air pollution constitutes one of the main environmental risk factors for public health, particularly in urban environments with limited real time monitoring infrastructure. Although low-cost IoT architectures have emerged as scalable alternatives to extend the spatial coverage of measurements, many implementations lack statistically validated risk classification models capable of translating the environmental data into information to service the public. This study presents the development and statistically validates in real time a risk index for air quality, implemented though a low-cost IoT architecture which integrates supervised artificial intelligence (AI) models deployed at the edge. The system was implemented for eight weeks the Universidad de Guayaquil campus, taking records of PM2.5, PM10 concentrations, temperature and humidity using calibrated sensors. The classification model based on Classification and Regression Trees reached a global accuracy greater which surpassed 90%, with 91.7% concordance to data retrieved from official stations for moderate conditions of PM2.5. K-fold method was used during the validation process and direct comparison with certified infrastructure. Additionally, user evaluation (n = 85) showed 82% adoption of preventive behavior after the implementation of proactive communication with a conversational bot. The results show that the integration of the presented low-cost IoT with validated risk models and user focused communication can generate reliable environmental intelligence and supports preventive decision-making in urban contexts with infrastructure constraints.
Some of the metrics are blocked by your
consent settings
A Multi-Camera Temporal Fusion for False Alarm Suppression in Edge-Based Weapon Surveillance
(Institute of Electrical and Electronics Engineers (IEEE), 2026)
Fierro Silva, Carlos Julio
;
Del-Valle-Soto, Carolina
;
Mostafa, Samih M.
;
Karim, Faten Khalid
;
Varela Aldas, José
Real-time weapon detection is a critical component of intelligent surveillance systems, particularly for perimeter monitoring applications on embedded edge platforms. However, reliable alarm generation remains challenging because false positives, temporal instability, and viewpoint inconsistencies can propagate through conventional multi-camera fusion strategies. To address these limitations, this work proposes a lightweight Adaptive Multi-Camera Temporal Fusion (ACTF) framework that combines confidence-aware evidence separation, temporal persistence, and short-window cross-camera validation at the decision level, thereby confirming detections without requiring additional neural-network inference. The framework was evaluated using a TensorRT-optimized YOLO26s detector in controlled dual-camera scenarios involving clear visibility, partial occlusion, visually ambiguous distractors, and challenging illumination. While logical OR fusion achieved higher recall, it also propagated erroneous detections; in contrast, ACTF completely suppressed the distractor-induced false alarms while maintaining competitive performance and sub-second confirmation under favorable conditions. The original NVIDIA Jetson Nano implementation achieved an average throughput of 2.4 camera-pair cycles per second, corresponding to low-rate online embedded operation, whereas an additional NVIDIA Jetson Xavier NX benchmark achieved an average of 10.2 camera-pair cycles per second. This is equivalent to 10.2 processed frames per second for each camera stream and 20.4 camera images per second in aggregate. These results support reactive real-time embedded operation on the Xavier NX platform and demonstrate that ACTF improves alarm reliability with negligible decision-level computational overhead. © 2013 IEEE.
Some of the metrics are blocked by your
consent settings
Using Predictive Modeling to Assess Social Media's Impact on Parental Well-Being: A Pilot Study
(Springer Nature Switzerland, 2026)
Tutillo, Jimena
;
Martinez, Danilo
;
Zapata, Mireya
Social media is part of daily life, facilitating communication, entertainment, and immediate access to information. However, it can negatively affect personal well-being and contribute to the development of addictive behavior. The objective of this paper is to examine how social media users in the economically productive age group of 36 to 56 years are affected by these platforms, and how machine learning (ML) and app development can be used to mitigate the impacts of these dopamine-stimulated behaviors. For this purpose, an empirical study using the Social Media Disorder Test showed that 37.5% of the sample had prolonged usage. To aid in the solution, an app was developed that uses machine learning (ML) to classify addiction levels into 8 categories across 3 social media apps. The app monitors usage in real time and issues alerts for excessive consumption patterns to reduce digital dependence. The results show real patterns of moderate addiction among users. It also seeks to raise awareness of social media use so they can set limits and reduce the impact of their use
Some of the metrics are blocked by your
consent settings
Real-Time Urban Animal Monitoring Using Transfer Learning-Based Object Detection on Web Platforms
(MDPI AG, 2026-05-13)
Fierro Silva, Carlos Julio
;
Sánchez, Carlos A.
;
Sánchez, Jorge S.
;
Del-Valle-Soto, Carolina
;
Velasco, Nancy
This study addresses the growing need for scalable solutions to monitor domestic and stray animals in urban environments. The objective is to develop and evaluate a real-time animal detection system using transfer learning and lightweight object detection models. The methodology includes the adaptation of a custom dataset with annotated images of cats and dogs under real-world conditions, followed by preprocessing, data augmentation, and model fine-tuning. Two architectures, SSD-MobileNet and YOLOv26s, were trained and evaluated using standard metrics such as precision, recall, F1-score, and mAP, as well as operational indicators like inference speed and system responsiveness. The best-performing model was integrated into a web-based platform with real-time detection, mobile access, and automated alerts. Results show that YOLOv26s outperforms SSD-MobileNet, achieving higher precision and recall while significantly reducing false positives and improving background discrimination. The system demonstrates near real-time performance suitable for monitoring applications and effective deployment across different input sources. The discussion findings highlight that integrating detection models with notification and visualization tools enhances practical applicability. Although SSD-MobileNet is suitable for low-resource environments, YOLOv26s provides a better balance between accuracy and reliability, making it more appropriate for real-world intelligent monitoring systems
Some of the metrics are blocked by your
consent settings
Rethinking criminal profiling through cognitive artificial intelligence
(Frontiers Media SA, 2026-05-08)
Buele, Jorge
;
Miranda-Toapanta, Davis
;
Rojas-Cañizares, David
Introduction Traditional criminal investigation often struggles to integrate dispersed and heterogeneous information, delaying the identification of serial or escalating patterns. Advances in artificial intelligence (AI) and cognitive computing offer data-driven approaches for cross-source correlation and temporal anomaly detection. Methods A focused narrative review of peer-reviewed literature on AI applications in forensic analysis, pattern detection, and investigative support was conducted using major multidisciplinary databases. Selected studies were synthesized into two analytical dimensions: evidence correlation and anomaly detection, and further examined through a retrospective case-based illustration. Results AI-based approaches support the linkage of low-level traces with higher-level events, enabling structured reconstruction and large-scale pattern identification. Machine learning models integrate heterogeneous data into operational representations, achieving high predictive performance under controlled conditions, often reported above 90% accuracy. Statistical and learning-based methods also detect temporal compression, behavioral drift, and cross-source anomalies, revealing patterns that may remain fragmented under manual analysis. Retrospective examination of historically fragmented cases highlights longitudinal regularities, including shrinking inter-event intervals, increasing severity, and the accumulation of weak signals that become meaningful when analyzed jointly. Discussion AI contributes primarily at a methodological level by enabling continuous integration and re-evaluation of investigative signals, supporting more data-informed and longitudinally grounded profiling practices. However, a gap persists between high performance in controlled settings and limited validation in real-world contexts. Future research should prioritize empirical benchmarking using operational datasets, the development of explainable and auditable systems, and governance frameworks ensuring transparent, accountable, and human-supervised deployment.

Centro de investigación en Mecatrónica y Sistemas Interactivos

Filters

Settings

Sort By

Results per page