Digital Manufacturing Data Readiness Audit

+ Full Text

Digital Manufacturing Data Readiness Audit Self-Assessment Tool

Digital Manufacturing Data Readiness Audit

Sight Machine works with global manufacturers looking to use production data to predict machine failure, optimize processes, and increase output. During our work, we’ve discovered that most of these efforts don’t fulfill expectations due to shortcomings in the condition of production data. In order for data to support digital manufacturing initiatives, it is critical to understand its readiness for integration with other production data sources. The ability to join, blend and integrate multiple sources of manufacturing data into digital twins of production processes, lines, plants, and parts is a foundational requirement for being a data driven manufacturer. Unfortunately, most manufacturers don’t understand the condition or readiness of their production data until after their project has begun.

Through numerous engagements, we’ve found that performing a comprehensive data audit is the most crucial step for ensuring a project is successful. The results of this audit allow manufacturers to properly scope project objectives, expectations, and timelines. Most importantly, by going through the audit process, manufacturers gain organizational alignment on what is required to enable the use of real-time production data. Performing an audit is the first step in building a digital manufacturing data strategy.

The Attributes That Impact Readiness At Sight Machine, we’ve developed a process for auditing production data to better understand its ability to deliver business impact. The Sight Machine Data Readiness Audit examines the attributes we’ve found to be most critical for integrating production data in real-time with other sources. These include

  1. Accessibility: The data can be accessed in a repeatable, automated method
  2. Format/ Protocol/ Schema: The data is in a consistent, readable format for ingestion into analytic applications
  3. Labeling: The data can be tied to physical sources involved in production
  4. Data Relatability: There is a systematic method for joining the data with other data streams
  5. Process Alignment: Resources are available for mapping data to processes and physical sources

To support real-time manufacturing analytics capabilities, a data source must be ready in all five data attributes. By assessing the readiness for each source, a manufacturer will be able to properly scope project timelines and budget.

info@sightmachine.com © 2018 Sight Machine

Digital Manufacturing Data Readiness Audit

Which Data Sources Should You Audit? The Data Readiness Audit is a valuable tool for analyzing the ability of individual data sources to support realtime manufacturing analytics capabilities. To properly analyze manufacturing processes, manufacturers need to integrate data from multiple sources involved in production. For production analysis, typically, these sources fall into 5 categories: Data Category Typical Characteristics Machine and Sensor Information

Time series information typically acquired from a Historian or SCADA system. Typical formats includes: § Timestamp (to the second or sub second) of the recording § Sensor name § Sensor reading

Product Information

Which type of product is being produced on a machine. This is typically captured through three methods:

  1. The product name / recipe is a tag on the machine that can be captured with machine and sensor information OR can be inferred through set points on the machine.
  2. From an ERP, MES, or scheduling system, if this is utilized, there is typically the need to understand business rules to associate this information with Machine data.
  3. Associating product type with a batch.

Downtime Classification

Associate downtime reasons with downtime events (detected with machine data). This typically comes from one of the following sources:

  1. Fault codes on the machine.
  2. A downtime tracking or maintenance system. This is typically associated with the machine data using time stamps
  3. Industrial PC or HMI connected to or near the machine

Batch Information

Information about batch number. This can be batches of material being input to a process (consumed batch) or the output batch number. This is typically done with timestamp association but some applications can also support business rules for association. Common data sources are: § Machine tags § MES or ERP systems

Defect or Scrap Information

Associating defective or scrap information to the batch or timestamp of machine production to be used as a categorical variable in analysis. This should be a predefined list of defect codes. In many organizations this may be a defect code hierarchy. Common data sources are: § MES or ERP Quality modules § LIMS Lab Information Management System § SPC software

Auditing your Data The Data Readiness Audit examines a data sources condition across each of the five attributes, allowing a manufacturer to determine if the data falls into one of three categories: Ready: The data source is ready to be used for real-time manufacturing analytics efforts Work to Do: Development resources can be applied to ready the data. This has implications for project timelines and budgets Not Ready: The data is not ready for real-time analytics efforts in it’s current form. This has implications on product scope and the use cases that can be addressed

info@sightmachine.com © 2018 Sight Machine

Worksheet No.

of

Date:

Digital Manufacturing Data Readiness Audit Self-Assessment Tool Ready

Work to Do

Auditor: Data Category: Data Location:

Not Ready

  1. Accessibility Data is accessible from the internet as it is generated (can be streamed or micro batched to the cloud) Action Needed: None! You’re ready for digital manufacturing

Data is captured and stored as it is generated but access to the file is a challenge Action Needed: Develop file transfer capabilities to move files to accessible location

Data can’t be retrieved in an automated fashion due to technical (e.g. data on an un- connected device) or policy issues (Security or cloud policies not in place) Action Needed: § Get IT buy-in to address policy issues § Evaluate infrastructure needs

Notes:

  1. Format / Protocol / Schema Data is defined and structured in a way that your modeling application can support Action Needed: None! You’re ready for digital manufacturing

Data is structured in a manner that is not currently supported Action Needed: Software development to transform data into a structure supportable by your application

No consistent, defined, or documented structure (e.g. manually entered data on a spreadsheet) Action Needed: Process change or system upgrades to enhance or remediate your data structure

Notes:

  1. Asset Tagging Data payload contains information on the physical sources (machine, product, component, line, etc.) Action Needed: None! You’re ready for digital manufacturing

The physical sources of the data are partially labeled or labeled manually

Physical source information for data is unavailable or not captured. Labels and metadata do not exists.

Action Needed: System development to assign physical sources and metadata

Action Needed: Investigate how to add asset IDs to your PLCs or capture process information

Notes:

  1. Data Relatability Timestamps or serialization exist enabling the data to be connected with other sources Action Needed: None! You’re ready for digital manufacturing

Timestamps can be inferred or data contains partial serial number information. Action Needed: Develop business logic to define relationships with other data sources

No consistency in timestamps or serialization of the data Action Needed: Add serialization equipment and/or synchronize system clocks with NTP across processes

Notes:

  1. Institutional Knowledge of Data Personnel are available that understand the data’s relationship to production processes Action Needed: None! You’re ready for digital manufacturing Notes:

info@sightmachine.com © 2018 Sight Machine

Personnel not available but documentation exists on processdata relationships Action Needed: Resource personnel to build institutional knowledge of data

Resources that understand the dataprocess relationships are not available and documentation does not exist Action Needed: Will require data forensics to examine data and build knowledge base