Grasping, Manipulation, and Contact Safety Performance of Robotic Systems

Summary

The Grasping and Manipulation project aims to develop and standardize performance metrics, test methods, and associated measurement tools that support the development of robotic systems capable of dexterous grasping and manipulation. This project focuses on advancing the state-of-the-art in robotic grasping and manipulation, enabling robots to effectively pick and manipulate objects in various applications.

The project will investigate new technologies, including advanced tactile sensing and AI-driven control strategies, to improve grasping and manipulation capabilities. It will also explore bi-manual manipulation systems, where two robots work together to manipulate a single object. By developing and standardizing test methods and measurement tools, this project will help to drive innovation and progress in the field of robotic grasping and manipulation.

Grasp strength test method with split cylinder artifact.

Assembly Task Board 1 in disassembled state.

Description

Credit: NIST

 

Objective
Provide performance metrics, test methods, and associated measurement tools to support next-generation robotic systems having human-like dexterity and control characteristics.

Technical Idea
The project will continue long-standing work and leadership within existing standards bodies, specifically ASTM International Committee F45 on Robotics, Automation, and Autonomous Systems, to develop new grasping and manipulation performance standards. Through close collaboration with industry partners, academia, and other stakeholders, the project will engage in the development and refinement of metrics, test methods, artifacts, and measurement tools that gauge the performance of robotic grasping and manipulation. The project will continue to respond the need for standards and metrics for emerging technologies to further enable the widespread adoption of robotic grasping and manipulation systems.

Research Plan

• Lead the development of a suite of robotic hand grasping and manipulation elemental performance metrics, test methods, and terminology within ASTM International sub-committee F45.05 on Grasping and Manipulation, with supporting testbed implementations.

• Lead the development of an ASTM F45.05 suite of standards supporting Grasp-type end-effector technology performance metrics, test methods and assembly task boards representing real-world tasks performed in the manufacturing industry.

• Develop measurement science needed for benchmarking robot assembly capabilities. Implement commercial-off-the-shelf (COTS) robotic technologies and use the developed measurement science to benchmark performance over many systems.

• Develop an automated benchmarking framework for robot assembly tasks, enabling users to easily document and compare their system's performance against others.

• Expand the catalog of manufacturing-based assembly tasks represented in the NIST ATB format, enabling broader evaluation and comparison of robotic assembly performance.

• Investigate the metrics and test methods needed for benchmarking robot capabilities intrinsic to multi-arm manipulation systems.

• Investigate the latest tactile sensing technologies for high dexterity robotic hands and develop testing methodologies to evaluate their intrinsic capabilities.

• Development of Robotic Perception and Manipulation Assets to Support Machine Learning Research in Manufacturing Applications including a synthetic Image Data Pipeline for Training Vision-Based Robot Systems and an accompanying image dataset of manufacturing parts.

---

Standards Development

• Established ASTM International sub-committee F45.05 on Grasping and Manipulation
• Assembly task boards
  • WK87214 Standard test method for robotic assembly task boards: Benchmarking performance
  • WK87213 Standard practice for robotic assembly task boards: Recording assembly test configuration
• Grasp type end-effectors
  • WK83863 Grasp-Type Robot End-Effectors: Grasp Strength Performance
  • WK86189 a Split Force Measurement Apparatus to Support Grasp-Type End-Effector Performance Tests
• ISO/TC 299 on Robotics
  • Assisted in the development of ISO TS 15066 Robots and Robotic Devices - Collaborative Robots within working group 3.
  • Assisted in the development of ISO PAS 5672:2023 Robotics – Collaborative Operations – Test Methods for Measuring forces and pressures in human-robot contacts within working group 8.

Datasets

• Designed and fabricated the publicly available NIST Manufacturing Objects and Assemblies Dataset (MOAD) data collection system
  (dataset and fabrication replication available at https://www.robot-manipulation.org/nist-moad)

Related Recent Publications

• Kimble, Kenneth, et al. "Performance measures to benchmark the grasping, manipulation, and assembly of deformable objects typical to manufacturing applications." Frontiers in Robotics and AI 9 (2022): 999348.
• Falco, Joseph, Karl Van Wyk, and Kenneth Kimble. "Advances in Robot Technology Supporting Low-Volume/High-Mix Small Part Assembly Operations." (2022).
• Kimble, Kenneth, et al. "Benchmarking protocols for evaluating small parts robotic assembly systems." IEEE robotics and automation letters 5.2 (2020): 883-889.
• Falco, Joe, et al. "Benchmarking protocols for evaluating grasp strength, grasp cycle time, finger strength, and finger repeatability of robot end-effectors." IEEE robotics and automation letters 5.2 (2020): 644-651.
• Case, Jennifer C., et al. "Towards the development of soft force and pressure sensors for robot safety applications." 2021 IEEE Sensors. IEEE, 2021.
• Sun, Yu, et al. "Research challenges and progress in robotic grasping and manipulation competitions." IEEE robotics and automation letters 7.2 (2021): 874-881.