NIST logo
*
Bookmark and Share

Smart Manufacturing Operations Management

Summary:

Efficient sensing and control is challenged by several realities at the factory floor, including product and process changes, fluctuating customer requirements, and even the profusion of proprietary software solutions. Although proprietary and home-grown solutions for manufacturing sensing and control have enabled significant efficiency improvements, these solutions are suboptimal. The focus of this project is to develop an integrated suite of standard methodologies and protocols. The suite targets both automated and human-assisted sensing and control leading to substantive improvements in manufacturing production efficiency and agility.

Description:

The project will focus on control and sensing at the Manufacturing Operation Management (MOM) level of the enterprise, though at the MOM level, “sensing” is commonly referred to “manufacturing metrics” or “Key Performance Indicators (KPI).” The current landscape of extant MOM level activities and information models, along with the methods used to accomplish the same, will be collected, analyzed, and documented. Following this activity, assessment of MOM level metrics and control, leading to a prioritization of the areas for project focus, will be performed. Project activities currently being worked on include standard methods and models for the effective selection and use of KPIs and the definition of information models for product quality measurement statistics.

Objective:

Deliver an integrated methodology and protocols to enable efficient networked and hierarchical sensing and control in smart manufacturing systems by FY 2018.


What is the new technical idea?

The technical idea is a suite of integrated standard methods and protocols, which will improve on sub-optimal efficiency solutions. Current efficiency solutions consist chiefly of non-standards based (i.e., proprietary) methods and protocols. Standard methods have been shown to be successful in a number of manufacturing1 domains, as have standard protocols2. However, there appears to be no standard, in the pattern of ISO 9001, for example, addressing factory system level sensing and control. These planned standard methods would not replace proprietary approaches. Rather, they will provide a vendor-neutral approach. Vendor-neutral approaches enable innovation and creativity, best-in-class solutions, "plug-and-play," and high information quality for solution vendors and manufacturers. What is the research plan? The research plan will focus on enabling more efficient sensing and control processes at the Manufacturing Execution Systems (MES) level of assembly-centric manufacturing. This focus will include 1) high level sensing information commonly called "performance metrics" or "key performance indicators" (KPI) and 2) the real-time correlation of product sensing information to process parameters, both for in-process and off-line sensing. In project out years, hierarchical and networked aspects of sensing and control will be addressed.

The first year will be dedicated to building NIST expertise through focused study, seeking new hires, forming collaborations, and documenting study results. All of this is to refine the milestones in the out years. We will study project-relevant best-in-class solutions and "reference approaches" to achieve efficient networked sensing and control. Such solutions and approaches should be those defined or currently used by metrics and efficiency consultants, thought leaders, and solution vendors. We will seek strong collaborative relationships with domain experts, and decide together what can and should be standardized. We will seek what industry, standards, and research individuals and organizations are best suited for standards development.

In years 2 and 3, the focus on the MES level will continue. This focus will include a standard model definition of sensing and control information critical for statistical process control and real-time correlation of this information with process parameters. In-process sensing is the goal , allowing feed-forward control. Much of this information is expected to be defined already in proprietary languages (digital and human) and will be collated in collaboration and with domain experts. Human-in-the-loop is anticipated. The work on standard methods and protocols will be done in partnership with key stakeholders . Stakeholders with hands-on experience will be sought.

In year 4, hierarchical and networked aspects will be examined and incorporated into the project. In year 5, method and model performance assurance will be addressed with the generation of validation and verification tools. In the last two years of the project it will be critical to incorporate classic sensing and control layers of the ISO 95 hierarchy. ISO 95 layers will include levels 0 and 1 with the higher levels (e.g., MES) in the complete hierarchy. Standard effective key performance indicator (KPI) sets and an accurate metrics maturity model, properly used, will contribute to the goals of this project, where KPIs are viewed aggregate sensing values.

Model and method validation are best done in both real and simulated environments. We plan to design and perform pilot tests in real/simulated manufacturing environments. Testing will be performed both at collaborator sites (industry and academia) and at NIST through existing and planned manufacturing testbeds. The project will focus on control and sensing at the Manufacturing Operation Management (MOM) level of the enterprise, though at the MOM level, “sensing” is commonly referred to “manufacturing metrics” or “Key Performance Indicators (KPI).” The current landscape of extant MOM level activities and information models, along with the methods used to accomplish the same, will be collected, analyzed, and documented. Following this activity, assessment of MOM level metrics and control, leading to a prioritization of the areas for project focus, will be performed. Project activities currently being worked on include standard methods and models for the effective selection and use of KPIs and the definition of information models for product quality measurement statistics.

1 e.g., ISO 9001, Baldrige Criteria, ASME's B89
2 e.g., Ethernet IP, DMSC's DMIS
3 MES is the production portion of Level 3 (Manufacturing Operations Management) of IEC 62264 (ISA-95)
4 The MESA and ISO TC184 SC5 organizations use this terminology (see ISO 22400).
5 "…quality is the combined effort of incoming material inspection, verification of each process step, and to the extent possible, verification of the success of a subassembly…" Randy Wire, Evana Automation
6 Manufacturers, efficiency solutions vendors, efficiency solutions consultants, industry organization, and standards organizations
7
John Jackiw and Dennis Brandl of the Manufacturing Enterprise Solutions Association (MESA) Metrics Working Group