Prognostics and Health Management for Reliable Operations in Smart Manufacturing (PHM4SM)

Summary

Develop and deploy measurement science to promote the implementation, verification, and validation of advanced monitoring, diagnostic, and prognostic technologies to increase reliability and decrease downtime in smart manufacturing systems.

Description

Smart manufacturing processes are becoming more complex, with increased integration of IIoT technologies, greater process reconfigurability to support product customization, and demands for higher precision. Maintaining these smart manufacturing systems through prognostics and health management has become more difficult, given the much larger range of influences on the processes. New PHM technologies are being developed, and verifying and validating these technologies will be important to reduce the risk of failure and increase confidence that systems will meet performance goals. This project will develop metrics and test methods for verification and validation of PHM technologies, and publish guidelines to support the integration of PHM for improving the reliability and resilience of smart manufacturing processes. 

WHAT IS THE TECHNICAL IDEA?
The technical idea is to develop measurement science to verify and validate the effectiveness of PHM technologies applied to manufacturing workcells. The results of this research will inform guidelines and standards that help manufacturers determine how best to instrument their workcells to provide prognostics information, which prognostics methodologies can be most effectively applied, and the improvements to reliability that can be expected. 

WHAT IS THE RESEARCH PLAN?
Through a combination of laboratory testbed research and pilot tests at industry sites, this project will develop measurement science to support the verification and validation of PHM technologies within manufacturing workcells, specifically for robotic arms and machine tools. The 2015 workshop on measurement science for PHM  noted several research opportunities in this area: a lack of rigorous measurement methods to enable efficient and effective data collection, the lack of validated models to predict the effectiveness of PHM prior to a system being put into practice, and inconsistent or insufficient standards that make it difficult to broadly apply PHM. 

The primary focus of the research will be to develop metrics for verifying and validating data collection methods. Prior work with robot arms led to a system for measuring end-of-arm positional accuracy from stereo cameras using targets robust to occlusion, for which a provisional patent is being obtained. This system will be enhanced with active illuminated smart targets that maximize observability, and test procedures that enable the identification of degradation sources down to the sensor and actuator level. Concurrently, tests will be developed that broaden this identification to sources throughout the workcell that influence performance on positioning tasks, using an artifact-based approach. For machine tools, the inertial measurement unit (IMU) sensor developed and tested in the NIST Shops will be extended from its current single-axis diagnostics capability to use for multi-axis prognostics. Following this, spindle performance measurement using sensors such as laser diodes and photodetectors will be integrated into a research platform that will inform the development of comprehensive test procedures for prognostics systems. 

Targeting the lack of validated models, the roadmap proposed developing methods and services to generate diagnostic and prognostic data sets for public use, including verification and validation. This would be supported by the development of specific test beds that would enable both the production of data and the necessary verification and validation. Addressing this challenge, the project will generate datasets from simulations, from equipment in the PHMC and Linear Axis Testbed laboratories, from the Smart Manufacturing Systems Testbed in the NIST Shops, and potentially from pilot sites. These datasets will be made available following the project’s Data Management Plan. 

The 2017 ASME/NIST workshop on Advanced Monitoring, Diagnostics, and Prognostics for Manufacturing Operations  identified seven standards priorities, three of which will be addressed by this project: standardized terminology, guidelines to determine what health data to capture and which collection strategies to use, and guidelines to determine which sensors should be deployed and how best to deploy them. These documents will be informed by the project’s research activities on metrics, by results from industry pilot projects, and by an ASME committee composed of members from industry and academia through which the documents will be published. 
 

Major Accomplishments

• Summary Report on the Industry Forum for Monitoring, Diagnostics, and Prognostics for Manufacturing Operations
• Summary Report on a Workshop on Advanced Monitoring, Diagnostics, and Prognostics for Manufacturing Operations
• Degradation Measurement of Robot Arm Position Accuracy Dataset
• Diagnostic and Prognostic Standards Gaps identified within Manufacturing Domains

• A Final Technical Status Report for Cost Effective Coordinated and Cooperative Robotics Enabled by Open TechnologiesPrepared by Southwest Research Institute

• Comprehensive Review of Diagnostic and Prognostic Capabilities and Best Practices For Manufacturing

• Roadmapping Workshop on Measurement Science for Prognostics and Health Management of Smart Manufacturing Systems
• Measurement Science for Prognostics and Health Management of Smart Manufacturing Systems Roadmapping Workshop Report
• Case Studies that Identify the Needs, Priorities, and Constraints of Small- to Medium-sized Manufacturers with respect to Performance Measurement, Condition Monitoring, Diagnosis, and Prognosis
• Definition of a Use Case that is Representative of Common Manufacturing Operations to Support the Assessment of PHM Technologies