Smart Technologies, Systems and Applications: 3rd International Conference, SmartTech-IC 2022, Cuenca, Ecuador, November 16–18, 2022, Revised Selected Papers

Preface
Organization
Contents
Smart Technologies
Demand Forecasting of Fast-Moving Consumer Goods by Deep Learning-Based Time Series Analysis
	1 Introduction
	2 Literature Review
		2.1 SARIMA-MLR Model
		2.2 Recurrent Neural Networks and LSTM Cells
		2.3 Time-Series Analysis in Demand Forecasting
	3 Materials and Methods
		3.1 Dataset Description
		3.2 SARIMA-MLR
		3.3 Recurrent Neural Networks
		3.4 Forecast Error Measures
	4 Results and Discussion
		4.1 SARIMA-MLR Equations
		4.2 Model Performance Comparison and Assessment
		4.3 Conclusions and Future Work
	References
Data Mining Application for the Generation of User Profiles in Serious Games Aimed at Attention and Memory Training
	1 Introduction
	2 Related Work
	3 Methodology
		3.1 Data Collection
		3.2 Data Understanding
		3.3 Data Preprocessing
		3.4 Modeling
		3.5 Pattern Analysis
	4 Results
		4.1 Performance Analysis
		4.2 Clustering Analysis
	5 Discussion
	6 Conclusions
	References
Automatic Class Extraction from Spanish Text of User Stories Using Natural Language Processing
	1 Introduction
		1.1 Related Work
	2 Material and Methods
		2.1 Dataset
		2.2 Natural Language Processing
	3 Experiments and Results
		3.1 Pattern Application to Extract Class Entities
		3.2 Processing Results
		3.3 Discussion
	4 Conclusions and Future Work
	References
A Tuning Approach Using Genetic Algorithms for Emergency Incidents Classification in Social Media in Ecuador
	1 Introduction
	2 Methods
		2.1 Problem Understanding
		2.2 Data Collection
		2.3 Data Understanding
		2.4 Modeling
		2.5 Testing
		2.6 Deploy
	3 Results and Discussions
		3.1 Performance
		3.2 Statistical Test
	4 Conclusion
	References
Software Hope Design for Children with ASD.
	1 Introduction
		1.1 Project Hope – Analytics
		1.2 User Centered Design
	2 Material and Method
		2.1 Work Plan
	3 Results
	4 Discussion
	5 Conclusions
	References
Performance Analysis of Emotion Recognition Prediction on Mobile Devices
	1 Introduction
	2 Related Works
	3 Methods
		3.1 Mobile Devices
		3.2 Data Base
		3.3 Algorithms
		3.4 System Description
		3.5 Metric
		3.6 System Performance Tests
		3.7 Usability
	4 Results
	5 Discussion and Conclusions
	References
Android Mobile Application for Cattle Body Condition Score Using Convolutional Neural Networks
	1 Introduction
		1.1 Problem Statement
		1.2 Related Works
	2 Materials and Methods
		2.1 Cattle Body Condition
		2.2 Data Set
		2.3 YOLO V3 Network
		2.4 MobileNet V2 Network
		2.5 VGG-16 Network
		2.6 Android Mobile App
	3 Results
		3.1 Data Set
		3.2 Yolo V3 Network
		3.3 MobileNet V2 Network
		3.4 VGG-16 Network
		3.5 Android Mobile App
	4 Discussion
	5 Conclusions and Future Work
	References
Automatic Selection of Reference Lines for Spectrometer Calibration with Recurrent Neural Networks
	1 Introduction
	2 Background
		2.1 Basic Spectrometer Principle
		2.2 3D-Printed Low-Cost Spectrometer
		2.3 Manual Calibration Process
	3 Methodology
		3.1 Object Detection
		3.2 Reference Lines Dataset Collection
		3.3 Recurrent Neural Networks - RNNs
	4 Results and Discussion
		4.1 Experiments
		4.2 Hyperparameter Tuning
		4.3 Discussion
	5 Conclusion
	References
Virtual Learning Environment for Children with Down Syndrome Through Leap Motion and Artificial Intelligence
	1 Introduction
	2 Methodology
		2.1 Design
		2.2 Creating the Virtual Environment and Configuring the Leap Motion Sensor
		2.3 Visual Feedback
		2.4 Data Acquisition and Processing
		2.5 Neural Network
	3 Results
		3.1 Leap Motion Recognition
		3.2 Obtaining Signals
		3.3 Number Recognition
		3.4 Final Results
	4 Discussion
	5 Conclusions
	References
Building a Knowledge Graph from Historical Newspapers: A Study Case in Ecuador
	1 Introduction
	2 Related Work
	3 Issues in Our Historical Newspaper Corpus
	4 Building a Knowledge Graph from Historical Newspapers
		4.1 Data Extraction and Storage
		4.2 OCR Process
		4.3 Cleaning and Pre-processing
		4.4 Named Entity Recognition
		4.5 Geospatial Processing
		4.6 Keywords Detection
		4.7 Knowledge Graph Generation
		4.8 Explotation
	5 Conclusions and Future Work
	References
Smart Systems
Colombian Sign Language Classification Based on Hands Pose and Machine Learning Techniques
	1 Introduction
	2 Materials and Methods
		2.1 Dataset
		2.2 Feature Extraction
		2.3 Machine Learning Techniques
	3 Results and Discussion
	4 Conclusions and Future Works
	References
Spanish Stylometric Features to Determine Gender and Profession of Ecuadorian Twitter Users
	1 Introduction
		1.1 Related Work
	2 Method
		2.1 Dataset
		2.2 Feature Extraction
		2.3 Training and Evaluation
		2.4 Prediction
	3 Results
	4 Discussion
	5 Conclusions
	References
Identification of Tropical Dry Forest Transformation from Soundscapes Using Supervised Learning
	1 Introduction
	2 Materials and Methods
		2.1 Site of Study
		2.2 Recordings
		2.3 Feature Extraction and Classification Approaches
		2.4 Performance Measures
	3 Results
	4 Conclusions
	References
Atmospheric Dispersion Prediction for Toxic Gas Clouds by Using Machine Learning Approaches
	1 Introduction
	2 Related Work
	3 Scenarios
	4 Methodology
	5 Testing and Results
	6 Conclusions
	References
Facial Recognition System with Liveness Detection Integrated into Odoo for Attendance Registration in a Software Laboratory
	1 Introduction
	2 Methodology
	3 Results
		3.1 Face Recognition with the Pre-trained Model Face-Recognition
		3.2 Development of the Liveness Detection Model
		3.3 Odoo Integration with Facial Recognition and Liveness Detection Module
	4 Discussion
	5 Conclusions
	References
Classification of Epileptic Seizures Based on CNN and Guided Back-Propagation for Interpretation Analysis
	1 Introduction
	2 Methodology
		2.1 Data-Base
		2.2 Pre-processing and Feature Extraction
		2.3 Neural Network Design
		2.4 Interpretability Analysis
	3 Results and Discussion
	4 Conclusions
	References
An Approach to the Presumptive Detection of Road Incidents in Cuenca, Ecuador Using the Data from the Social Media Twitter and Spanish Natural Language Processing
	1 Introduction
	2 Related Works
	3 The Presumptive Detection of Road Incidents Application in Cuenca - Ecuador
		3.1 Stage 1. Twitter Data Collection
		3.2 Stage 2. Tweet Classification
		3.3 Stage 3. Integration of NLP Models with the Mobile Application
	4 Results
		4.1 Accuracy Testing
		4.2 User Acceptance Testing
	5 Conclusions
	References
Financial Credit Risk Measurement Using a Binary Classification Model
	1 Introduction
	2 Research Design
		2.1 Population and Sample
		2.2 Theoretical Foundation
	3 Predictive Model Development
	4 Results
		4.1 Evaluation of the Models
		4.2 Results Validation
	5 Discussion
	6 Conclusions and Future Work
	References
FPGA-Based Four-Band Multispectral Camera Prototype for Precision Agriculture Applications
	1 Introduction
	2 Multispectral Camera Development
		2.1 Hardware Development
		2.2 Software Development
	3 Image Registration
	4 Results and Discussion
		4.1 Image Registration
		4.2 NDVI Vegetation Index
	5 Conclusions
	References
Small-Scale Precision Agriculture: A Case Study Applied to a Seed Germination Chamber
	1 Introduction
	2 Literature Review
	3 Methodology and Materials
	4 Development
	5 Results
		5.1 Control System
		5.2 Development of Seedlings
	6 Discussion
	7 Conclusion
	References
Diagnosis and Degree of Evolution in a Keratoconus-Type Corneal Ectasia from Image Processing
	1 Introduction
	2 Related Work
	3 Methodology
		3.1 Scheimpflug Camera Images
		3.2 Edge-Sobel Image Filter and Thresholding
		3.3 Mathematical Analysis Berlin/Ambrosio
		3.4 Affectation Degree Classification Algorithm
	4 Results
		4.1 Variation in Berlin/Ambrosio Curves
		4.2 Parameters and Processing Image
		4.3 Affectation Classification Algorithm
		4.4 Performance Evaluation of CNN
	5 Discussion
	6 Conclusion
	References
Finite Element Analysis of Uncovered Coplanar Interdigital Electrodes for Capacitive Sensor Applications
	1 Introduction
	2 Theory
		2.1 The Model of Interdigital Electrodes Capacitors
		2.2 The Model of Interdigital Electrodes Capacitors Considering the Electrode Thickness
	3 Methodology
	4 Results
		4.1 Simulations of the Samples with a Variable Electrode Width and Substrate Thickness
		4.2 Simulations of the Samples with a Variable Electrode Width and Electrode Thickness
	5 Conclusion
	References
Smart Trends and Applications
Algorithm for Medical Diagnostic Support Using Machine and Deep Learning for Depressive Disorder Based on Electroencephalogram Readings
	1 Introduction
		1.1 Depressive Disorder
		1.2 Electroencephalogram (EEG) for Detection of Depressive Disorder
		1.3 State of the Art
	2 Materials
	3 Methods
		3.1 Machine Learning Techniques
		3.2 Classification
		3.3 Deep Learning Techniques
	4 Results
		4.1 Result of Classification Based on Each Power EEG Bands
		4.2 Results of Classification Based on Combining Power EEG Bands
		4.3 Results of Deep Recurrent Neural Network ChronoNet
	5 Discussion
	6 Conclusions
	References
Effects of a Robotic Lower-Limb Orthosis on Gait Based on the Analysis of sEMG Signals and Kinematic Data: Preliminary Results
	1 Background
	2 Materials and Methods
		2.1 Subjects
		2.2 Robotic Lower-Limb Orthosis ALLOR
		2.3 Signal Acquisition
		2.4 Protocol
		2.5 Data Processing and Analysis
	3 Results
		3.1 Myoelectric Activity
		3.2 Kinematic Analysis
		3.3 Technology Satisfaction
	4 Discussion and Conclusions
	References
Design and Evaluation of a Prototype of Dual Channel Electrical Stimulator: Application in the Assessment of Current Perception Thresholds to Multiple Stimuli
	1 Introduction
	2 Materials and Methods
		2.1 Design
		2.2 Functional Tests
		2.3 Experimental Protocol
	3 Results
		3.1 Functional Evaluation
		3.2 Experimental Evaluation
	4 Discussion and Conclusions
	References
System for Postural Stability Assessment Based on an Inertial Sensors Network
	1 Introduction
	2 Methodology
		2.1 Experimental Protocol
		2.2 Calibration and Alignment IMUs
		2.3 Centre of Mass (COM)
	3 Results and Discussions
	4 Conclusion
	References
Perspectives on Climate Change Until 2030
	1 Introduction
	2 Methodology
		2.1 Processing in the Frequency Domain
		2.2 Neural Networks
	3 Experimentation
		3.1 Data Set Selection
		3.2 Data Set Analysis
		3.3 Forecasted Trend
	4 Conclusions
	References
Virtual Power Plants as a Management Strategy to Improve Resilience in Electrical Distribution Networks
	1 Introduction
	2 Methodology
	3 The Proposed Scheme for Coordinating Energy Dispatch Using VPP in the Case of a Disruptive Event in the Network
	4 Test Scenario: Scenario with Disruptive Event: Energy Dispatch in the Network Under Atypical Operating Conditions When No Conventional Distribution Company is Operating
	5 Results Obtained
	6 Conclusions
	References
Design of an Integrated System for the Measurement and Monitoring of Electrical Parameters Under ISO 50001: A Case Study
	1 Introduction
	2 Methodology
		2.1 Hardware Design
		2.2 Structure of the Network
		2.3 Overall Design and Communication of the Monitoring System
		2.4 Modbus Communication Protocol
		2.5 Electrical Data Acquisition with PLC S7–1200
		2.6 Communication Protocol S7
		2.7 Hosting Implementation
		2.8 Web Application Development Methodology
	3 Results
		3.1 Web Application Graphical Interface
		3.2 Analysis of Results
	4 Conclusions
	References
Aerodynamic Analysis Based on Materials and Shapes of Different Types of Wind Turbine Blades for Wind Energy Production Using Fast and MATLAB Simulink
	1 Introduction
	2 Materials and Methods
		2.1 Selection of the Aerodynamic Variables to Be Studied
		2.2 Coupling of FAST and MatLab Simulink
		2.3 Obtaining the Blades to Be Tested
		2.4 Software Simulation
	3 Results
		3.1 Results of the Analysis of Blades with the Use of Aluminum 2014T6
		3.2 Results of the Analysis of Blades with the Use of Balsa Wood
		3.3 Results of the Analysis of Blades with the Use of Composite Material
	4 Conclusions
	References
Numerical Comparison Using CFD of R134a and R600a Condensation in Helical and Spiral Exchangers
	1 Introduction
	2 Materials and Methods
	3 Results
	4 Conclusions
	References
CIEMS: A Review Towards a Collaborative Industrial Emergency Management Information System
	1 Introduction
	2 Methodology
	3 Background Work
		3.1 Academic Proposals
		3.2 Comparison of Existing Information Systems
		3.3 Mobile Applications for Emergency Management
	4 Architecture of the Collaborative Management System
		4.1 Description
		4.2 Architecture
		4.3 Services
	5 Conclusions and Future Works
	References
PID and Model Predictive Control (MPC) Strategies Design for a Virtual Distributed Solar Collector Field
	1 Introduction
	2 Description and Mathematical Model of the Virtual Distributed Solar Collector Field
		2.1 Description of the Virtual Distributed Solar Collector Field
		2.2 Mathematical Model of the Virtual Distributed Solar Collector Field
	3 Virtualization Methodology
	4 Design of Control Algorithms
		4.1 PID Control Strategy Design
		4.2 Model Predictive Control MPC Strategy Design
	5 Results
		5.1 Analysis of the Virtual Distributed Solar Collector Field Implemented
		5.2 Analysis of the Control Strategies Implemented in the Virtual Distributed Solar Collector Field
	6 Conclusions
	References
Speed Controller by Neural Networks Trained by Invasive Weeds for a DC Motor
	1 Introduction
	2 Methodology
		2.1 Data Acquisition
		2.2 Control Topology
		2.3 Tuning by MatLab PID Tuner
		2.4 Neuro Controller Development
		2.5 Tuning by IWO Algorithm
	3 Results
		3.1 Discrete-Time Plant Responses for Speed Control of a DC Motor
		3.2 Statistical Results Using the Wilcoxon Test for PID and Back-Propagation Tuned Controllers
		3.3 Statistical Results Using the Wilcoxon Test for Back-Propagation and IWO Tuned Controllers
	4 Discussion
	5 Conclusions
	References
Abiotic Maize Stress Detection Using Hyperspectral Signatures and Band Selection
	1 Introduction
	2 Materials and Methods
		2.1 Database
		2.2 Band Selection Techniques
		2.3 Classification Techniques
	3 Results
		3.1 Classification Results
		3.2 Best Selected Bands
		3.3 Discussion
	4 Conclusion
	References
Optimal Energy Dispatch Analysis Using the Inclusion of Virtual Power Plants Based on Dynamic Power Flows
	1 Introduction
	2 Virtual Power Plant (VPP)
		2.1 Economic Dispatch
	3 Problem Statement
		3.1 Dynamic Optimal Dispatch Model Based on FOPAC
	4 Analysis of Results
		4.1 Study Case
	5 Conclusions
	References
Routing Protocols Implementation for WSN Using Optimal Dimensioning
	1 Introduction
	2 Wireless Sensor Networks
		2.1 Dimensioning for WSN
	3 ZigBee-Based WSN Implementation
		3.1 Routing Protocols for WSN Evaluated
	4 Analysis of Results
	5 Conclusions
	References
Author Index