Artificial Intelligence for Robotics and Autonomous Systems Applications

Preface
Contents
Efficient Machine Learning of Mobile Robotic Systems Based on Convolutional Neural Networks
1 Introduction
2 Problem Analysis and Formulation
3 Efficient Deep Learning for Robotics—Related Work
3.1 Efficient Deep Learning Models for Object Detection in Robotic Applications
3.2 Efficient Deep Learning Models for Semantic Segmentation in Robotic Applications
4 The Proposed Models of Efficient CNNs for Semantic Segmentation Implemented on Jetson Nano
5 Obstacle Avoidance Algorithm for Mobile Robots Based on Semantic Segmentation
6 Experimental Results
7 Discussion of Results
8 Conclusion
Appendix A
References
UAV Path Planning Based on Deep Reinforcement Learning
1 Introduction
1.1 Research Background and Significance
1.2 Research Status
1.3 The Main Research Content and Chapter Arrangement of this Chapter
2 Deep Learning and Reinforcement Learning
2.1 Comparison of Supervised Learning, Unsupervised Learning and Reinforcement Learning
2.2 Deep Learning Methods
2.3 Reinforcement Learning Methods
2.4 DQN Algorithm
3 Design of Improved DQN Algorithm Combined with Artificial Potential Field
3.1 Network Structure Design
3.2 State Space Design
3.3 Action Space Design
3.4 Reward Function Design
4 Simulation Experiment and Result Analysis
4.1 Reinforcement Learning Path Planning Training and Testing
4.2 Training and Results
4.3 Comparative Analysis of Improved DQN and Traditional DQN Algorithms
5 Conclusions
References
Drone Shadow Cloud: A New Concept to Protect Individuals from Danger Sun Exposure in GCC Countries
1 Introduction
2 Related Work
2.1 Overview
2.2 Proposal
3 Proposed Method
3.1 Design of Mechanical Structure
3.2 Shade Fabric
3.3 Selecting a Umbrella Flight Controller
3.4 Flying Umbrella Power Calculation
4 Experiment Phase
5 Results and Discussion
6 Conclusion
References
Accurate Estimation of 3D-Repetitive-Trajectories using Kalman Filter, Machine Learning and Curve-Fitting Method for High-speed Target Interception
1 Introduction
2 Related Work
3 Vision Based Target Position Estimation
3.1 Computation of Centre of Instantaneous Curvature of Target Trajectory
3.2 EKF Formulation
3.3 Future State Prediction
4 Mathematical Formulation for Curve Fitting Method
4.1 Classification of Curves
4.2 Least-Squares Curve Fitting in 2D
4.3 Least-Squares Curve Fitting of Any Shape in 3D
5 Interception Strategy
6 Results
6.1 Simulation Experiments
6.2 Hardware Experiments
7 Discussions
8 Conclusions
References
Robotics and Artificial Intelligence in the Nuclear Industry: From Teleoperation to Cyber Physical Systems
1 Introduction
1.1 Background
1.2 Motivation
1.3 Problem Statement
1.4 Recent Technological Advances–Industry 4.0
1.5 Chapter Outlines and Contributions
2 Nuclear Decommissioning Processes
2.1 Characterisation
2.2 Decontamination
2.3 Dismantling and Demolition
2.4 Waste Management
3 Current Practice in Nuclear Decommissioning Research
3.1 Assisted Teleoperation and Manipulation in Nuclear Decommissioning
3.2 Robot-Assisted Glovebox Teleoperation
3.3 Post-processing of Nuclear Waste
3.4 Modular and Cooperative Robotic Platforms
3.5 Unmanned Radiation-Monitoring Systems
4 Towards an Autonomous Nuclear Decommissioning Process
4.1 Different Levels of Autonomy
4.2 The Cyber Physical System Architecture
4.3 Enabling Technologies
5 A Cyber Physical Nuclear Robotic System
5.1 Software Architectures
5.2 Autonomous Multi-robot Systems
5.3 Control System Design
5.4 Motion Planning Algorithms
5.5 Vision and Perception
5.6 Digital Twins in Nuclear Environment
6 Conclusions
References
Deep Learning and Robotics, Surgical Robot Applications
1 Introduction
2 Related Work
3 Machine Learning and Surgical Robot
4 Robotics and Deep Learning
5 Surgical Robots and Deep Learning
6 Current Innovation in Surgical Robotics
7 Limitation of Surgical Robot
8 Future Direction of Surgical Robot
9 Discussion
10 Conclusions
References
Deep Reinforcement Learning for Autonomous Mobile Robot Navigation
1 Introduction
2 Antecedents
2.1 Control Theory, Linear Control, and Mechatronics
2.2 Non-linear Control
2.3 Classical Robotics
2.4 Probabilistic Robotics
2.5 Introduction of Back-Propagation for Feed-Forward Neural Networks
2.6 Deep Reinforcement Learning
3 Background: Autonomous Mobile Robot Navigation and Machine Learning
3.1 Requirements
3.2 Review of RL in AMR
3.3 Introduction of Convolutional Neural Networks
3.4 Advanced AMR: Introduction to DRL (Deep Reinforcement Learning) Approach
3.5 Application Requirements
4 Deep Reinforcement Learning Methods
4.1 Continuous Control
4.2 A Simple Implementation of Q-learning
4.3 Dueling Double DQN
4.4 Actor-Critic Learning
4.5 Learning Autonomous Mobile Robotics with Proximal Policy Optimization
4.6 Multi-Agent Deep Reinforcement Learning
4.7 Fusion Method
4.8 Hybrid Method
4.9 Hierarchical Framework
5 Design Methodology
5.1 Benchmarking
6 Teaching
7 Discussion
8 Conclusions
References
Event Vision for Autonomous Off-Road Navigation
1 Introduction
2 Related Work
2.1 Off-Road Navigation
2.2 Neuromorphic Vision
2.3 Key Features of Event Cameras
2.4 Data Registration
2.5 Feature Detection
2.6 Algorithmic Compatibility
3 Event-Based Vision Navigation
3.1 Vision and Ranging Sensor Fusion
3.2 Stereo Event Vision
3.3 Monocular Depth Estimation from Events
4 Proposed End-to-End Navigation Model
4.1 Event Preprocessing
4.2 Depth Estimation Branch
4.3 Steering Prediction Branch
4.4 Desert Driving Dataset
5 System Implementation
5.1 Deep Learning Acceleration
5.2 Memristive Neuromorphic Computing
6 Results and Discussion
7 Conclusions
References
Multi-armed Bandit Approach for Task Scheduling of a Fixed-Base Robot in the Warehouse
1 Introduction
2 Related Work
3 Problem Formulation
3.1 Problem Framework and Description
3.2 Motion Planning
4 Methodology
4.1 Multi-armed Bandit Formulation
4.2 Task Scheduling Based on Time Synchronization
5 Results and Analysis
5.1 Simulation Results
5.2 Discussion
6 Conclusion
References
Machine Learning and Deep Learning Approaches for Robotics Applications
1 Introduction
2 Autonomous Versus Automatic Robots
3 Robotics Applications
3.1 Computer Vision
3.2 Computer Vision
3.3 Learning Through Imitation
3.4 Self-supervised Learning
3.5 Assistive and Medical Technologies
3.6 Multi-agent Learning
4 Extreme Learning Machines Methods for Robotics
5 Machine Learning for Soft Robotics
6 ML-Based Robotics Applications
6.1 Robotics Recommendation Systems Using ML
6.2 Nano-Health Applications Based on Machine Learning
6.3 Localizations Based on ML Applications
6.4 Control of Dynamic Traffic Robots
7 Robotics Applications Challenges
8 Conclusions
References
A Review on Deep Learning on UAV Monitoring Systems for Agricultural Applications
1 Introduction
2 Proposed Methodology
2.1 An Overview of Deep Learning Strategies used in Agriculture
3 Findings on Applications of Deep Learning Models in Plant Monitoring
3.1 Pest Infiltration
3.2 Plant Growth
3.3 Fruit Conditions
3.4 Weed Invasion
3.5 Crop Disease Monitoring
4 Findings on Applications of Deep Learning Models in Animal Monitoring
4.1 Animal Population
5 Discussion and Comparison of Deep Learning Strategies in Agricultural Applications
6 Conclusions
References
Navigation and Trajectory Planning Techniques for Unmanned Aerial Vehicles Swarm
1 Introduction
2 UAV Technical Background
2.1 UAV Architecture
2.2 UAV Swarm Current State
2.3 UAV Swarm Advantages
2.4 UAV Swarm Applications
3 Swarm Communication and Control System Architectures
3.1 Centralized Communication Architecture
3.2 Decentralized Communication Architecture
4 Navigation and Path Planning for UAV Swarm
4.1 UAVs Network Communication and Path Planning Architecture
4.2 Trajectory Planning for UAVs Navigation Classifications
4.3 Route Planning Challenges
5 Classical Techniques for UAV Swarm Navigation and Path Planning
5.1 Roadmap Approach (RA)
5.2 Cell Decomposition (CD)
5.3 Artificial Potential Field (APF)
6 Reactive Approaches for UAV Swarm Navigation and Path Planning
6.1 Genetic Algorithm (GA)
6.2 Neural Network (NN)
6.3 Firefly Algorithm (FA)
6.4 Ant Colony Optimization (ACO)
6.5 Cuckoo Search (CS)
6.6 Particle Swarm Optimization (PSO)
6.7 Bacterial Foraging Optimization (BFO)
6.8 Artificial Bee Colony (ABC)
6.9 Adaptive Artificial Fish Swarm Algorithm (AFSA)
7 Conclusions
References
Intelligent Control System for Hybrid Electric Vehicle with Autonomous Charging
1 Introduction
2 Preliminaries
2.1 Hybrid Vehicle and Pure Electric Vehicle
2.2 Hybrid Vehicle Architecture
3 The Architecture of Electric Vehicles
3.1 Battery Technologies
3.2 Super-Capacitors
3.3 The Electric Motor
4 Electric Vehicles Charging
4.1 Types of Classic Chargers
4.2 Autonomous Charger
5 The Mathematical Model for the Autonomous Charging System
5.1 Inductive Power Transfer Model
5.2 Photovoltaic Generator Model
6 Simulation Results and Discussion
6.1 Fuzzy Logic Algorithms
6.2 Power Delivered by the Charging System
6.3 Power Distribution and SOC Evolution
7 Conclusion
References
Advanced Sensor Systems for Robotics and Autonomous Vehicles
1 Introduction
1.1 Automatic Driving Application
1.2 Railway Monitoring Application
2 Related Works
3 Types of Sensors for Various Applications
3.1 Efficient Road Monitoring
3.2 Efficient Railway Monitoring Monitoring
4 Conclusion
References
Four Wheeled Humanoid Second-Order Cascade Control of Holonomic Trajectories
1 Introduction
2 Related Work
3 Robot Motion Model
4 Observer Models
5 Omnidirectional Cascade Controller
6 Results Analysis and Discussion
7 Conclusions
References