IoT Edge Solutions for Cognitive Buildings

Preface
Contents
1 COGITO: A Platform for Developing Cognitive Environments
	1.1 Introduction
	1.2 The COGITO Platform
		1.2.1 An Overview of the Platform
		1.2.2 Developing an Application Over the COGITO Platform
	1.3 Equipment Deployment
	1.4 Cognitive Applications for Indoor Environments
		1.4.1 Thermal Comfort
		1.4.2 Occupancy Forecast
		1.4.3 Air Quality
		1.4.4 Smart Meeting Room
	1.5 Cognitive Applications for Outdoor Environments
		1.5.1 Smart Parking
		1.5.2 Monitoring Weather Conditions
	1.6 Conclusion
	References
2 Cloud, Fog, and Edge Computing for IoT-Enabled Cognitive Buildings
	2.1 Introduction
		2.1.1 Background
	2.2 Cloud Computing
		2.2.1 Benefits of Cloud Computing
		2.2.2 Cloud Computing Characteristics
		2.2.3 Cloud Services Models
		2.2.4 Cloud Deployment Models
		2.2.5 Cloud Computing for Smart Building
	2.3 Edge and Fog Computing
		2.3.1 Edge Computing Characteristics
		2.3.2 Fog Computing
			2.3.2.1 Fog Computing Characteristics
			2.3.2.2 Fog Computing Services and Deployment Models
		2.3.3 Fog Computing Versus Edge Computing
		2.3.4 Edge Computing Versus Cloud Computing
		2.3.5 Edge and Fog Computing for Smart Buildings
	2.4 IoT-Enabled Smart Buildings
		2.4.1 IoT-Enabled Smart Building Model
		2.4.2 IoT-Enabled Smart Building Components
		2.4.3 IoT-Enabled Smart Building Design Framework
		2.4.4 Sample IoT-Enabled Smart Building Scenarios
			2.4.4.1 HVAC System
			2.4.4.2 Health Monitoring System
			2.4.4.3 Remote Monitoring System
	2.5 Conclusion
	References
3 Edge Caching in IoT Smart Environments: Benefits, Challenges, and Research Perspectives Toward 6G
	3.1 Introduction
	3.2 Content Distribution in Smart Environments
		3.2.1 Peculiarities of IoT Contents and Devices
		3.2.2 Benefits of Edge Caching
	3.3 Conventional Edge Caching Designs
	3.4 Disruptive Pervasive Edge Caching Solutions
		3.4.1 NDN in a Nutshell
		3.4.2 Popularity-Driven Approaches
		3.4.3 Energy-Driven Solutions
		3.4.4 Freshness-Driven Caching
			3.4.4.1 Freshness-Based Only Solutions
			3.4.4.2 Multi-Criteria Approaches
	3.5 IoT Data Caching at the Edge Toward 6G
		3.5.1 NDN and SDN Interplay
		3.5.2 Joint Computing and Caching
		3.5.3 AI-Based In-Network Caching Strategies
	3.6 Conclusions
	References
4 Needs Analysis, Protection, and Regulation of the Rights of Individuals and Communities for Urban and Residential Comfort in Cognitive Buildings
	4.1 Introduction
	4.2 New Technologies for Outdoor and Indoor Well-Being: The Legal Framework
	4.3 Multiscalar Analysis of the City of Matera and the Demonstrators of the COGITO Project
	4.4 An “Ideal” Technology: Reflections from Ethnographic Research
	4.5 Need Assessment in the Design of a Cognitive System
	4.6 Effectiveness of the GDPR in the Data Protection System Transformed by New Technologies
	References
5 Real Case Studies Toward IoT-Based Cognitive Environments
	5.1 Introduction
	5.2 Wireless Sensor Networks and MANs for CoIoTEs
		5.2.1 Wireless Sensor Networks
		5.2.2 MAN
		5.2.3 Applications and Design Challenges
	5.3 Implementation of the Smart Street Network in the City of Cosenza
		5.3.1 Description of the Communication Backbone
		5.3.2 Implementation of the Final Solution
		5.3.3 The Devices Involved in the Realization of the Smart Street
	5.4 A VPN ``Hub and Spoke\'\' for Secure Interconnection of Geographically Distributed Sensor Networks
		5.4.1  VPN Topology Overview
		5.4.2 A Secure Sockets Layer VPN
		5.4.3 Realization of the Peripheral Hub Nodes and the Master Server
	5.5 Design and Implementation of an Intelligent Video Conferencing System for CoIoTEs
		5.5.1 The Jitsi Video Conferencing System
		5.5.2 A Web-Based File Manager
		5.5.3 The Email Processing Component
	5.6 Conclusions
	References
6 Audio Analysis for Enhancing Security in Cognitive Environments Through AI on the Edge
	6.1 Introduction
	6.2 Approaches for Audio Recording
		6.2.1 Array of Microphones
		6.2.2 Recording Devices
	6.3 Understanding Audio Recordings
	6.4 AI in Audio Analysis
	6.5 AI and Algorithms for Sample Normalization and Audio Understanding
	6.6 Privacy Implications on Sensitive Data: Defining Minimum Information Content
	6.7 Analyzing Hardware Devices
	6.8 Reacting Based on the Information: Actuators
	6.9 A Complete Implementation
	6.10 Case Study
		6.10.1 Free and Restricted Access Room
		6.10.2 Residential Apartment
	6.11 Further Improvements/Conclusions
	References
7 Aggregate Programming for Customized Building Management and Users Preference Implementation
	7.1 Introduction
	7.2 Description of the Brescia Use Case
		7.2.1 User Preferences and Feedback Collection
		7.2.2 Sensor Integration Through IoT Paradigm
	7.3 Aggregate Programming
	7.4 Aggregate Programming for the Brescia Use Case
		7.4.1 Users Localization
		7.4.2 Porting Aggregate Programming to Embedded Systems
		7.4.3 An AP Case Study Using RTLS
	7.5 Simulation
	7.6 Conclusions
	References
8 IoT Control-Based Solar Shadings: Advanced Operating Strategy to Optimize Energy Savings and Visual Comfort
	8.1 Introduction
	8.2 Materials and Method
		8.2.1 Sensors and Actuators
		8.2.2 Operating Control Strategy for Venetian Blinds
			8.2.2.1 Absence of Occupants
			8.2.2.2 Presence of Occupants
	8.3 Analysis of Results
		8.3.1 Simulation Environment
		8.3.2 Evaluation of Thermal Gains on an Hourly Basis
			8.3.2.1 LED System
			8.3.2.2 Fluorescent Lamps
		8.3.3 Evaluation of Thermal Gains on a Monthly Basis
		8.3.4 Evaluation of Annual Energy and Economic Savings
			8.3.4.1 Energy Savings
			8.3.4.2 Economic Savings
	8.4 Conclusions
	References
9 Room Occupancy Prediction Leveraging LSTM: An Approach for Cognitive and Self-Adapting Buildings
	9.1 Introduction
	9.2 Related Work
	9.3 An Approach for Room Occupancy Prediction for Cognitive and Self-Adapting Building
		9.3.1 Software Architecture
			9.3.1.1 Components
			9.3.1.2 Devices and Virtual Objects
			9.3.1.3 Application Agents
		9.3.2 Definition of the Prediction Tasks
		9.3.3 Data Pre-processing
			9.3.3.1 Transformation Approach for Non-homogeneous Time Series
		9.3.4 Networks Training
			9.3.4.1 LSTM Neural Network
		9.3.5 Training
	9.4 Experimental Results
		9.4.1 Dataset
			9.4.1.1 Dataset A: Occupancy Detection Dataset
			9.4.1.2 Dataset B: Experimental Dataset
		9.4.2 Evaluation Metrics
		9.4.3 Imbalanced Classification Techniques
			9.4.3.1 Focal Loss
			9.4.3.2 Weight Balancing
		9.4.4 Results
			9.4.4.1 Task 1: Occupancy Detection
			9.4.4.2 Task 2: Occupancy Prediction
	9.5 Conclusion and Future Work
	References
10 Edge Intelligence Against COVID-19: A Smart University Campus Case Study
	10.1 Introduction
	10.2 Background and Enabling Technologies
		10.2.1 ACOSO-Meth
		10.2.2 Uppaal
		10.2.3 DHT11
		10.2.4 Arduino Uno
		10.2.5 QR Code
		10.2.6 Raspberry Pi
		10.2.7 Node-RED
		10.2.8 MQTT (Message Queue Telemetry Transport)
		10.2.9 Long Short-Term Memory (LSTM)
		10.2.10 Docker
		10.2.11 DigitalOcean
	10.3 Related Works
		10.3.1 Monitoring at the End-Device Layer
		10.3.2 Monitoring at the Edge Layer
		10.3.3 Monitoring at the Cloud Layer
	10.4 Project Development
		10.4.1 Analysis Phase
		10.4.2 Design Phase
		10.4.3 Verification and Validation
		10.4.4 Implementation Phase
		10.4.5 Deployment and Orchestration
	10.5 Conclusions
	References
11 Structural Health Monitoring in Cognitive Buildings
	11.1 Introduction
	11.2 Structural Monitoring Techniques
	11.3 Cognitive Buildings
	11.4 Case Study
	11.5 Conclusions and Future Activities
	References
12 Development of Indoor Smart Environments Leveraging the Internet of Things and Artificial Intelligence: A Case Study
	12.1 Introduction
	12.2 Related Work
	12.3 Smart Management of Indoor Spaces
	12.4 Smart Meeting Room Application Components
		12.4.1 Smart Objects
		12.4.2 Software Components
	12.5 Management of the Conference System in Indoor Environments
		12.5.1 Management of the Booking of the Smart Meeting Room
		12.5.2 Event Management in the Pre and Start Phases
		12.5.3 Event Management
	12.6 Conclusion
	References
13 Human-Centered Reinforcement Learning for Lighting and Blind Control in Cognitive Buildings
	13.1 Introduction
	13.2 Reinforcement Learning in Control Systems
	13.3 A Human-Centered RL with a Satisfaction-Based Visual Comfort Model
	13.4 An RL Model for the Management of the Visual Comfort
		13.4.1 The State Variables
		13.4.2 The Decision Variables
		13.4.3 The Reward Function
		13.4.4 Q-Learning
	13.5 Case Study
	13.6 Conclusions
	References
14 Intelligent Load Scheduling in Cognitive Buildings: A Use Case
	14.1 Introduction
	14.2 Basic Concepts
		14.2.1 The COGITO Platform
		14.2.2 Reinforcement Learning
		14.2.3 Markov Decision Process
		14.2.4 The Load Scheduling
	14.3 Integration Between the COGITO Platform and the Omnia Energia Equipment
	14.4 The Case Study
		14.4.1 The Case Study Equipment
		14.4.2 The Functional Perspective
		14.4.3 The Underpinning Software Infrastructure
		14.4.4 Customization of the Omnia Meter
		14.4.5 The Case Study Dashboard
	14.5 Conclusion
	References
15 Cognitive Systems for Energy Efficiency and Thermal Comfort in Smart Buildings
	15.1 Introduction
	15.2 Related Work
	15.3 A DRL Model for the Management of Indoor Environments
		15.3.1 Thermal Model
		15.3.2 Behavioral Model
		15.3.3 Objective Function and Reward
	15.4 Experimental Results
	15.5 Conclusions
	References
Index