Advances on Robotic Item Picking: Applications in Warehousing & E-Commerce Fulfillment

Preface
Contents
Team CVAP\'s Mobile Picking System at the Amazon Picking Challenge 2015
	1 Introduction
	2 Platform
	3 Strategy
	4 Behavior Trees
		4.1 Semantic
		4.2 BTs in APC
	5 Shelf Localization and Base Movement
	6 Vision
		6.1 Simtrack
		6.2 Volumetric Reasoning
	7 Picking and Placing
	8 Conclusion
	References
Team UAlberta: Amazon Picking Challenge Lessons
	1 Introduction
	2 Strategy
		2.1 Pulling–Pushing Mechanism
	3 System Description
		3.1 Pulling Routine
		3.2 Centering Routine
		3.3 Grasping Routine
	4 Conclusions
	References
A Soft Robotics Approach to Autonomous Warehouse Picking
	1 Introduction
	2 Robot Design
		2.1 Qbmoves
		2.2 Pisa/IIT SoftHand
	3 Perception
		3.1 Point Cloud Database
		3.2 Perception Algorithm
	4 Motion Planning and Control
		4.1 Motion Planner
		4.2 Grasping Strategies
	5 Conclusions and Lesson Learned
	References
Top-Down Approach to Object Identification in a Cluttered Environment Using Single RGB-D Image and Combinational Classifier for Team Nanyang\'s Robotic Picker at the Amazon Picking Challenge 2015
	1 Introduction
	2 The Picking Problem
	3 The Robotic Picking System
	4 Top-Down Identification Method
		4.1 Feature Extraction
		4.2 Probabilistic Modeling
	5 Results and Discussion
		5.1 Image Identification
		5.2 Competition Performance
	6 Conclusion and Future Work
	References
Team KTH\'s Picking Solution for the Amazon Picking Challenge 2016
	1 Introduction
	2 System Overview and High-Level Logic
		2.1 Stowing Strategy
		2.2 Picking Strategy
	3 Perception
		3.1 Shelf Pose Estimation and Tote Calibration
		3.2 Object Detection
	4 Motion Planning, Pick and Place
		4.1 Motion Planning
			4.1.1 Outside-Shelf Motion Planning
			4.1.2 Inside-Shelf Motion Planning
		4.2 Pick and Place
	5 Conclusions
	References
End-to-End Learning of Object Grasp Poses in the Amazon Robotics Challenge
	1 Introduction
	2 Related Work
		2.1 Amazon Robotics Challenge
		2.2 Object Grasp Pose Synthesis
	3 System Overview
		3.1 System Setup
		3.2 Process Flow
	4 Vision Subsystem
		4.1 Model Overview
		4.2 Data Collection
		4.3 Pre-training by CG Images
		4.4 Training
		4.5 Inference
	5 Results
	6 Future Work
	7 Conclusion
	References
A Systems Engineering Analysis of Robot Motion for Team Delft\'s APC Winner 2016
	1 Introduction
	2 A Framework to Analyze an Autonomous Robot Design
		2.1 Functional Analysis Under the ISE&PPOOA Method
		2.2 Levels of Robot Automation
	3 Motion Subsystem Design
		3.1 Motion Requirements
		3.2 Robot Manipulator
		3.3 Motion Software Module Design
		3.4 Offline Coarse Motions
		3.5 Grasping and Manipulation: Fine Motions
			3.5.1 Collision Avoidance
			3.5.2 Generation of the Complete Grasp Plan
		3.6 Trajectory Execution
			3.6.1 Trajectory Stitching
			3.6.2 Input/Output Synchronization
	4 Discussion and Concluding Remarks
		4.1 Concluding Remarks
	References
Standing on Giant\'s Shoulders: Newcomer\'s Experience from the Amazon Robotics Challenge 2017
	1 Introduction
	2 Technical Challenges
	3 Approach
		3.1 Past Competitions
		3.2 Design Philosophy
		3.3 Development Strategy
	4 Proposed Solution
		4.1 Suction Tool
		4.2 Storage System
	5 Conclusion
	Appendix
	References
Team C2M: Two Cooperative Robots for Picking and Stowing in Amazon Picking Challenge 2016
	1 Introduction
	2 Two Cooperative Industrial Robots
		2.1 Robot System
		2.2 Features
	3 Vision Strategy
		3.1 Grasping Position Based Object Recognition by CNN
		3.2 Picking Strategy
		3.3 Stowing Strategy
	4 Experiments
		4.1 Dataset
		4.2 Accuracy for Object Recognition
		4.3 Accuracy for Grasping Position Detection
		4.4 Computational Time
	5 Conclusion
	References
Description of IITK-TCS System for ARC 2017
	1 Introduction
	2 Object Recognition Algorithms
	3 Custom Gripper Design
	4 Grasping Algorithm
	5 Robot-Camera Calibration
	6 Motion Planning with Obstacle Avoidance
	7 Software Architecture for Implementation
	8 System Performance
	9 Conclusions
	References
Designing Cartman: A Cartesian Manipulator for the Amazon Robotics Challenge 2017
	1 Introduction
	2 Background
	3 Cartman: System Overview and Design
	4 Mechanical Design
		4.1 Specifications
		4.2 Mechanical Components
		4.3 Software and Electrical Components
		4.4 Multi-Modal End-Effector
	5 Perception System
		5.1 Perception Pipeline
		5.2 Perception Hardware
		5.3 Semantic Segmentation
			5.3.1 Fast Data Collection and Quick Item Learning
			5.3.2 Implementation and Training Details
		5.4 Active Perception
			5.4.1 Multi-View
			5.4.2 Item Reclassification
		5.5 Grasp Synthesis
	6 Design Philosophy
		6.1 End-to-End Testing
		6.2 Modularity
		6.3 Rapid Iteration
	7 System Performance
		7.1 Amazon Robotics Challenge Finals
		7.2 Long-Term Testing
		7.3 Error Detection and Recovery
	8 Conclusions
	References
Index