This book aims at sharing knowledge about the technological opportunities and the main research challenges regarding robotics for logistics in supermarkets and retail stores, from the perspectives of the end-users, logistic companies, technology providers, and robotic researchers. The authors have been involved into the H2020 project Robotics Enabling Fully Integrated Logistics Lines for Supermarkets (REFILLS), aimed at improving logistics in supermarkets thanks to mobile robotic systems in close and smart collaboration with humans. The readers will find a comprehensive analysis of the main logistic processes in retail stores with possible robotized solutions, involving mechanical design, perception, and control. These technologies have been validated in realistic environments, and some of them have been tested into real supermarkets. The book is intended for a broad academic and industrial readership, including operators in the field of logistics, distribution, and retail.
Author(s): Luigi Villani, Ciro Natale, Michael Beetz, Bruno Siciliano
Series: Springer Tracts in Advanced Robotics, 148
Publisher: Springer
Year: 2022
Language: English
Pages: 181
City: Cham
Series Editor’s Foreword
Preface
Contents
Contributors
1 Robot-Assisted Instore Operations in Grocery Retailing
1.1 Introduction
1.2 Characterization of Grocery Supply Networks
1.3 Instore Logistics Systems and Challenges
1.3.1 Instore Logistics Processes
1.3.2 Instore Logistics Costs
1.3.3 Specifics of Instore Logistics Systems
1.3.4 Past Developments and Status Quo of Instore Logistics
1.3.5 Challenges for Instore Automation Approaches
1.4 Automation Approaches to Support Instore Logistics
1.4.1 Scenario 1—Store Monitoring
1.4.2 Scenario 2—Collaborative Instore Logistics
1.4.3 Scenario 3—Autonomous Shelf Filling
1.5 Managerial Challenges of Using Professional Service Robots in Retail Stores
1.6 Conclusion
References
2 Robots Collecting Data: Modelling Stores
2.1 Introduction
2.2 Conceptual Framework and Computational Problem
2.3 SemDT Data and Knowledge Model
2.4 Virtual Simulation Environment of the Semantic Digital Twin
2.5 Semantic Mapping
2.5.1 Layout Identification
2.5.2 Store Monitoring
2.6 Example Store Monitoring System
2.6.1 Mobile Robotic Platform
2.6.2 Perception
2.7 Conclusion
References
3 Robots Collecting Data: Robust Identification of Products
3.1 Introduction
3.2 Related Work
3.3 Method
3.3.1 Network Architecture
3.3.2 Triplet Loss
3.3.3 Triplet Mining
3.3.4 Training
3.4 Datasets
3.4.1 Synthetic Dataset
3.4.2 Real World Dataset
3.4.3 Test Dataset
3.4.4 Omniglot
3.5 Experiments
3.5.1 Evaluation Metric
3.5.2 Selection of Embedding Size
3.5.3 Results
3.5.4 Performance on Neural Compute Stick 2
3.6 Conclusion
References
4 Robots Working in the Backroom: Depalletization of Mixed-Case Pallets
4.1 Introduction
4.2 Gripper
4.2.1 Mechanical Structure
4.2.2 Sensors and Actuators
4.2.3 Grasping Procedure
4.3 Detection, Recognition and Localization
4.3.1 The Cases Database
4.3.2 The Detection Module
4.3.3 The Geometrical Module
4.4 Integrated Cell
4.4.1 Cell Design
4.4.2 Executive System
4.4.3 Selection and Grasping
4.5 Experiments
4.5.1 Quantitative Analysis
4.5.2 Real Supermarket Scenario
4.6 Conclusion
References
5 Robots Helping Humans: Collaborative Shelf Refilling
5.1 Introduction
5.2 Handling Module
5.3 Methodology
5.3.1 Procedure
5.3.2 Ergonomic Indices
5.4 Simulations and Ergonomic Handling Module (EHM)
5.4.1 Simulation Setup
5.4.2 Data Collection
5.4.3 Data Processing
5.5 Experiments
5.5.1 Experimental Setup and Protocol
5.5.2 Experimental Data Collection
5.5.3 Validation of the Virtual Simulations
5.5.4 Validation of the Ergonomic Indices
5.6 Conclusions
References
6 Robotic Clerks: Autonomous Shelf Refilling
6.1 Introduction
6.2 Software Architecture for Autonomous Shelf Refilling
6.3 The Sensorized Gripper
6.3.1 Tactile Sensor Electronics
6.3.2 Tactile Sensor Mechanics
6.3.3 Sensor Calibration
6.4 Reactive Control System
6.4.1 Grasp Control
6.4.2 Visual Servoing
6.5 Plan Language for In-Store Manipulation Activities
6.5.1 Library of Generalized Fetch and Deliver Plans for In-Store Manipulation
6.5.2 Integration of the Reactive Controller
6.6 Experience-Based Learning and Optimization of Manipulation Activities
6.6.1 Learning Pipeline
6.6.2 Generating Episodic Memories within a Simulator
6.6.3 Learning from Episodic Memories
6.7 Experiments
6.7.1 Replenishment Experiments
6.7.2 Evaluation of Learning Pipeline
6.8 Conclusions
References
