4. Enterprise Manufacturing Visibility (EMV)

This section provides a detailed description of the visibility tools in Sight Machine’s Enterprise Manufacturing Visibility (EMV) applications. It contains the following sections:

Exploring Raw Data

EMV includes several visualization tools that allow you to explore your raw data, namely:

Raw Data Monitoring

You can view raw data in real time as it streams into the AI Data Pipeline.

On the Raw Data tab, begin by clicking  Raw Data Monitoring.

On the main Raw Data Monitoring screen, you can choose your options on the left, and then update to monitor streaming raw data.

The Raw Data Monitoring options include:

  1. Asset: You can select an asset from your raw data.
  2. Raw Fields: You can select up to five of the associated raw fields being captured for the selected asset.
  3. Start Date: Set to Live by default, which displays a real-time view of the data. You can also select a date and/or time in the past, which will stream data starting from that point.
  4. Frequency: You can select the streaming frequency of the visualization. This setting lets you control the pace at which data is visualized, irrespective of the pace at which it is collected. For example, Fast means you will see less data on-screen, but it will update more frequently. Slow means you will see more data on-screen, but it will update less frequently.
  5. Update: Click this button to start or restart the visualization.

The following is an example of raw data monitoring output:

Raw Data Streams

You can view a list of streams that have come in from FactoryTX, broken down by Asset and Stream Type.

On the Raw Data tab, begin by clicking  Raw Data Streams.

The following is an example of raw data streams output:

Raw Data Table

You can browse raw data or download the results into a CSV file to work with it in another application.

On the Raw Data tab, begin by clicking  Raw Data Table.

On the main Raw Data Table screen, choose your options on the left, and then update.

The Raw Data Table options include:

  1. Asset: You can select an asset from your raw data.
  2. Stream Types: You can select one or more stream types. All are selected by default, and you can remove stream types to apply a filter.
  3. Date Range: You can click the date range selector, and then select the range that you want to inspect.
  4. Update: Click this button to generate a flattened table of your raw data.
    NOTE: If the results contain more than 10,000 records, the interface will only display the first 10,000.

After you update, you will see the raw data table.

You can also click the download button in the upper-right to generate a CSV file of the results.

Raw Data Visualization

You can generate time series and distribution charts of your raw data.

On the Raw Data tab, begin by clicking  Raw Data Visualization.

On the main Raw Data Visualization screen, you can choose your options on the left, and then update.

The following is an example of the raw data visualization landing page:

The Raw Data Visualization options include:

  1. Asset: You can select an asset from your raw data.
  2. Stream Types: You can select one or more stream types. All are selected by default, and you can remove stream types to apply a filter.
  3. Raw Field: You can select one raw field to visualize over time.
    NOTE: You can later select a second field, but are limited to one in this window.
  4. Date Range: You can click the date range selector, and then select the range that you want to display.
  5. Show by: You can change the chart from time series to distribution (i.e., how often the variable you selected occurs at certain values).
  6. Update: Click this button to generate the chart.
    NOTE: If the results contain more than 10,000 records, the interface will only display the first 10,000.

The following is an example of raw data visualization output:

You can show a second field on the chart. To do this, in the upper-right, click the Y axis button, make a field selection, and then click  Update. The chart will now display the two color-coded fields.

Exploring Contextualized Data

The following tools allow you to visualize and work with the contextualized data models created by the Sight Machine AI Data Pipeline:

Data Tables

To create a new Data Table, on the Data Tab, begin by clicking  Create New Table.

On the new table screen, you can choose your options on the left, and then generate a table.

There are two tabs at the top of the sidebar. The first tab contains all of the menus for selecting data. The second tab allows you to filter the results.

The table options include:

  1. Model: You can select from among the contextualized data models. With EMV, this includes Cycles, Downtimes, OEE, and optionally Defects. With EMA, there is an additional option for Parts/Lots.
  2. Assets: You can select specific machines, sorting by machine type or location. On the Cycles, Downtimes, and Defects models, you can only select machines of the same type. On the OEE model, you can select machines across different types.
    NOTE: When the Parts/Lots model is selected, the Asset selector is replaced by a Part Type/Lot Type selector.
  3. Date Range: You can click the date range selector, and then select the range that you want to inspect. A relative date range will be selected by default. You can change this to a historical date range by unchecking the box marked Relative Date.
  4. Data Fields: You can select any number of parameters to display.

To apply filters to the data, click the Filter tab, click Add Filter, and then select a data field. For categorical fields, you can select the value or set of values (and toggle between equal to and not equal to). For continuous fields, you can apply one or more numeric conditions (e.g., range, greater than, etc.).

After you update, you can save your table by clicking the save icon at top of the page. Provide a name for the table, and then click  Save. If you make additional changes, be sure to click Save when you are finished.

Use the icons in the upper-right to share a link to the table or download the results as a CSV.

Data Visualization

The Data Visualization tool allows you to visualize contextualized data in a variety of ways, such as the trend of a variable over time, the distribution of the values of a variable, or the comparison of two variables. This simple point and click interface extends the power of sophisticated data querying across your organization.

On the Analysis tab, begin by clicking  Data Visualization.

On the main data visualization screen, you can choose your options on the left, and then update.

At the top of the sidebar on Data Visualization, you will see three tabs. The first tab contains all of the standard menus for generating a chart. The second tab allows you to create and apply one or multiple filters to the resulting data. The third tab contains additional advanced options for selecting and visualizing data.

The options on the first tab include:

  1. Model: You can select from among the contextualized data models. With EMV, this includes Cycles, Downtimes, OEE, and optionally Defects. With EMA, there is an additional option for Parts/Lots.
  2. Assets: You can select specific machines, sorting by machine type or location. On the Cycles, Downtimes, and Defects models, you can only select machines of the same type. On the OEE model, you can select machines across different types.
    NOTE: When the Parts/Lots model is selected, the Asset selector is replaced by a Part Type/Lot Type selector.
  3. Date Range: Unlike traditional reporting tools, the Sight Machine platform is constantly pulling in real-time data. Live Data is selected by default, which will continue to update the chart with live data. You can uncheck this option to visualize a static date range.
  4. Y Axis: You can set the Y axis on your chart. The options are dependent on the model and asset(s) that you selected. For example, the Cycles model includes some defaults (such as cycle count, cycle time, shift) as well all continuous data fields from the selected machine type.
  5. X Axis: You can set the X axis on your chart. The options are dependent on the model and asset(s) that you selected. For example, the Cycles model includes some defaults (such as time, cycle time, shift) and all of the continuous and categorical data fields from the selected machine type.
  6. Chart Type: You can display the data as a bar, line, scatter, heatmap, or box.
  7. Update: Click this button to generate your data visualization chart.

To apply filters to the data, click the  Filter tab, click Add Filter, and then select a data field. For categorical fields, you can select the value or set of values (and toggle between equal to and not equal to). For continuous fields, you can apply one or more numeric conditions (e.g., range, greater than, etc.)

The Advanced tab allows you to compare your data by an additional variable, change the aggregation function (if any), and set axis limits.

There are up to three options:

  • Compare By: Compare machine or factory.

  • Aggregates: Aggregate the Y axis variable as an average per cycle, or minimum, maximum, or total of all cycles. This is hidden when there is no aggregation applied (with a scatter plot, for example).

  • Y Axis Limits: Set fixed limits on the Y axis. By default, the interface will adjust the limits automatically.

Click the  Update button to display or recalculate.

If you selected a time series chart, the X axis will display the time interval. Click this to view the list of options and change the interval.

There are also a number of options available to you in the upper-right. To display the data as a table, click the  Table button. Click again to toggle back to Chart.

To share this visualization with others, click the  Share button.

To download the results as a CSV, click the  Download button.

KPI Visibility

The KPI Visibility tool allows you to visualize key performance indicators (KPIs), including OEE, the components of OEE (Availability, Performance, and Quality), and other models, such as downtimes and defects. These KPIs are built from machine-level and system-level data, and are calculated in a consistent manner for all facilities feeding data into the Sight Machine platform. This enables your organization to have a standardized method for comparing performance and other KPIs across your machines, lines, and facilities.

On the Analysis tab, begin by clicking  KPI Visibility.

On the main KPI Visibility screen, there is no Update button. Any time that you change a selection, the chart updates automatically.


The KPI Visibility options include: 

  1. Assets: You can select which assets you want to view, sorting by asset type or location.
  2. Date Range: Unlike traditional reporting tools, the Sight Machine platform is constantly pulling in real-time data. Live Data is selected by default, which will continue to update the chart with live data. You can uncheck this option to visualize a static date range.
    NOTE: You can only select dates, not times, as KPIs are calculated on a per-Shift basis.
  3. Y Axis: You can select the KPI that you want to display on the Y axis.
  4. X Axis: You can set the X axis on your chart. The available options will depend on your Y axis selection.
  5. Compare By: You can compare the data by asset, asset type, or factory.
  6. Display Type: You can display the data as a bar, line, pie, pareto, or number display.

After you make your changes, the chart updates automatically.

If you select Time on the X axis, after the chart loads, you will see Time in Days at the bottom of the chart area. You can then click  Days and adjust the time interval of the chart.

Working with Dashboards

This section provides an overview of the dashboard functionality in Sight Machine’s Enterprise Manufacturing Visibility (EMV) application. Topics include:

All dashboards are built off of the same machine-level data, which is constantly updated in real time, so every user in the organization has access to the most current information and metrics.

You can select any of the dashboards on the Dashboards tab, including:

  • Out-of-the-box dashboards
  • Public dashboards created by others in your organization and shared with you
  • Private dashboards that you have created

Viewing Global Operations

The Global Operations View is an out-of-the-box dashboard that provides visibility into output from manufacturing facilities across your enterprise, allowing corporate executives and production analysts to see unit production trends by facility. The Sight Machine platform pulls this output data in real time directly from machine-level data at each facility. Traditionally, this type of data is pulled manually and rolled up on a periodic basis, such as daily or weekly.

To select an individual facility in order to get more detail, you can click a facility on the map or select one using the scroll bar along the bottom.

If you do not select a facility, the Global Operations View will automatically scroll through each facility, one at a time. In this mode, it can be the ideal dashboard to display overall enterprise output and activity via kiosks or lobby displays throughout your organization.

Using Out-of-the-Box Dashboards for Specific Users

There are two standard dashboards that provide insight for specific users within your enterprise:

Using the Plant Manager Dashboard

This dashboard provides performance metrics such as OEE and a breakout of availability, quality, and performance for the entire manufacturing facility, as well as aggregate production in real time for the facility. This information allows plant leaders to track and compare key performance metrics by shift and machine type.

The following is an example of a Plant Manager Dashboard:

Using the Operator Dashboard

This dashboard provides a view of the real-time availability and performance of an operator’s machine. Typically, a screen with this dashboard would be placed next to the workstation of a machine operator to display a snapshot of key performance metrics such as overall production, unplanned downtimes, and average cycle times. Most importantly, it lets the operator see how the machine’s performance compares to other similar machines within the facility or across the enterprise. 

The following is an example of an Operator Dashboard:

Using the Dashboard Builder

Sight Machine’s dashboard builder allows you to create personalized combinations of charts for yourself or, by changing the settings to public, for others in your organization.

To use the dashboard builder:

  1. On the Dashboards tab, click Create New Dashboard.
  2. In the blank dashboard work table that opens, provide a meaningful name. For example, Lasercut Example Dashboard.
  3. You can make any private dashboard public and available to others in your organization by changing the permissions settings at the top of the dashboard as follows:
    • Public: Visible to anyone in your organization. However, the creator is the only one who can edit.
    • Private: Only visible to the user who created it. This is the default setting.
  4. Click the Add Widget icon to add a widget to the dashboard.
  5. In the Add a new widget window that appears in the upper-right corner, select any of the following:
    • Data Visualization: This allows you to chart a data field against time or against another data field.
    • KPI: This allows you to select a performance indicator such as quality and chart it over time or shift.
    • Raw Data Monitoring: This allows you to monitor raw data fields in real time.
  6. NOTE: For the EMA product offering, an additional menu item appears in this window for Statistical Process Control. For more information on using this option, see Using the Statistical Process Control Tool.

  1. After you select a widget type, an Edit Widget screen containing the selected tool will open. Make your selections on the left, and then click Update.
    For more information of the specific widget types, see the Data Tables, Data Visualization, or KPI Visibility section above.
  2. The name field is pre-populated with text based on the variables that you selected. You can edit the name to provide more information.

  3. To save the widget to your dashboard, click the Save button.

Editing Dashboard Widgets

After you save your newly created widget, it appears on your dashboard. 

In the upper-right of any widget, click the three dot menu to do any of the following:

  • Edit: This will take you back to the same screen where you created the widget.
  • Make a Copy: This creates a copy of the widget, and automatically takes you into the edit screen.
  • Delete: This removes the dashboard.

Copying a widget is a great time saver for creating many similar widgets with slightly different selections. For example, you can build a widget that visualizes data from one machine, and then create a copy for each other machine of the same type.

To further customize your dashboard, you can drag and drop the widgets to create a layout that works for you. To change the size of a widget, click and drag the lower-right corner to resize it to increase visibility or allow you to fit more widgets. This is especially useful if there are certain machines or data relationships that you compare regularly.

Within any dashboard, click the three dot menu to do any of the following:

  • Create a new widget.
  • Delete the dashboard.



After you delete the dashboard, you see a list of all the public dashboards to which you have access, as well as any private dashboards. Under Dashboard Name, you can click any dashboard to view it, or in the upper-right, click the plus (+) button to create a new one. 

You can also access this Dashboards list directly from the Dashboards tab. After you begin creating dashboards, an option appears on the Dashboards tab that allows you to view the full list (in case it does not fit in the menu).

Using the Statistical Process Control Tool

The Statistical Process Control (SPC) tool allows you to monitor process stability to improve product quality and optimize manufacturing processes. The SPC tool plays a critical role in production because stability monitoring allows you to:

  • Increase output while minimizing scrap, waste, rework, and defects.
  • Prevent process errors that lead to defects, letting you consistently deliver the highest quality product.
  • Deliver actionable information to personnel, minimizing the response time to correct process variation.
  • Support compliance and regulatory efforts through cost-effective recording of part/batch process variability.

The SPC tool provides:

  • Contextualized models that allow for segmentation of machine/line data based on the production of different parts/SKUs.
  • User-defined specification limits and calculated control limits that identify when processes are not acting predictably.
  • Nelson rule application that detects non-random conditions in the data stream.
  • Descriptive statistics such as min, mean, max, and CPK/CP values.

It is important to remember that the parameter data being monitored in the SPC tool is built on the data modeled by Sight Machine’s AI Data Platform. The AI Data Pipeline uses machine learning and AI to summarize parameter readings for each cycle, eliminating noise from higher granularity readings that may be too detailed for an exploratory analysis and ignoring variability that happens outside the cycle, such as during planned downtime, that could interfere with the analysis.

On the Analysis tab, begin by clicking  Statistical Process Control.

You can choose your options on the left to generate your chart.

The Statistical Process Control options include:

  1. Model: You can analyze only cycles at this time.
  2. Assets: You can select an asset, or multiple assets of the same type, to monitor.
  3. Date Range: Unlike traditional reporting tools, the Sight Machine platform is constantly pulling in real-time data. Live Data is selected by default, which will continue to update the chart with live data. You can uncheck option this to visualize a static date range.
  4. Data Field: You can select a parameter to monitor.
  5. Spec Limits: You can set specification limits for your manufacturing process.
  6. Chart Type: You can select the chart type most relevant to your process. You have a choice between I-MR, x-bar R, and x-bar S. For the x-bar R, and x-bar S options, you can also set a Subgroup Size, although it defaults to Auto.

To apply filters to the data, click the Filter tab, click Add Filter, and then select a data field. For categorical fields, you can select the value or set of values (and toggle between equal to and not equal to). For continuous fields, you can apply one or more numeric conditions (e.g., range, greater than, etc.)

Tool Output:

  • The SPC tool produces a histogram of data points post aggregation displayed on the left.
  • In the center is a time series chart with Nelson rule violations highlighted in red.
  • The time series chart includes dotted lines for control limits and specification limits.
  • The visualization also includes a data table summarizing information such as CPK and CP values.

Other Options:

  • Next to Nelson Rules, the blue information icon provides you with a definition of the Nelson rule codes. The following is an example of Nelson rules definition from the Statistical Process Control tool:

  • In the upper-right, there are several icons:
    • The N icon lets you toggle on and off specific Nelson rule violations from the chart.
    • The chart icon allows you to toggle the chart and table views.
    • The share icon lets you share a link to the specific analysis.

Operationalizing SPC:

  • Sight Machine can enable alerts to notify you when parameters go out of specification or control limits. Contact your Sight Machine engagement team or partner for more information on this capability.
  • The SPC charts can be incorporated into dashboards to enable machine operators or plant managers to monitor parameter stability. For more information, see Using the Dashboard Builder in Enterprise Manufacturing Visibility (EMV).

EMV Demonstration: Applying Insights in Your Organization

This section provides a walk-through of a hypothetical situation to show how different members of an organization would use Sight Machine’s Enterprise Manufacturing Visibility (EMV) application to identify an issue, investigate its causes, and drive continuous improvement efforts at a manufacturing facility. Topics include:

Identifying Issues

In this example, start by accessing the Global Operations View, which is available at the top of the Dashboards tab.

Using the Global Operations View, a corporate manufacturing leader can see a high-level view of output from different facilities across the enterprise.

In this hypothetical, the fusion machines at your facilities are experiencing variable throughput. Some facilities are producing more units of output per fusion machine than others. You are able to identify that the Harbin facility is having the lowest output per machine across the board and is a target for process improvement.

The Harbin facility manager has a dashboard on the Sight Machine platform specific to their role. The dashboard contains high-level KPIs, such as a real-time tracking of output and other OEE levers, specific to the Harbin facility.

The Sight Machine platform integrates and contextualizes data across all your machines globally. In this case, your Harbin Manager Dashboard shows fusion machine performance from all other facilities across the globe.

You know that fusion machine output is of concern, so by quickly looking at your dashboard, you can investigate potential issues. You can see that quality and performance for your fusion machines is somewhat similar, but availability is vastly different between your facility at Harbin and other fusion facilities.

In this case, the plant manager or a process engineer would investigate this availability issue. You need to know if your machines at Harbin are going down more frequently or if the length of their downtime is longer.

Investigating Causes

To investigate further, you can use Sight Machine’s Contextualized Data Visibility Toolkit (sometimes referred to as the Data Visualization tool), which is available on the Analysis tab.

Hypothesis 1: Are Downtime Durations Affecting Machine Availability?

You can check the downtime duration at your facilities to see if that is the factor driving your poor level of availability. Next to Assets, you select all 10 of the Fusion machines across all facilities.

Next to Model, you select Downtimes, because that is the focus of your investigation.

Next to X Axis, you select Downtime Duration to see the appropriate distribution.

The resulting chart gives you a view of how long it takes to recover from downtimes for all of your fusion machines across every facility.

NOTE: Remember, this data is pulled directly from machine data. No one had to run a special report, and all the information was calculated in the same manner. The Sight Machine platform acts as the single source of truth.

In the resulting chart, you can see if there was a discrepancy in the amount of time it takes to recover a machine at the Harbin facility vs. your other facilities.

What you discover is that it takes, on average, somewhere between a minute and a half to two minutes to remedy downtime issues. But you also see that it is fairly similar across your facilities, so it does not appear that downtime duration is the cause of your differences in output.

Hypothesis 2: Are the Number of Downtimes Affecting Machine Availability?

Your next step is to check to see if the number of downtimes is the issue. Are the machines at Harbin going down more often than fusion machines at other locations?

To investigate this, next to X Axis, you change Downtime Duration to Time instead.

HINT: You can type the data field name in the search bar to avoid scrolling through the list.

After clicking Update, in the resulting chart, you can now compare downtime count by facility.

You see that the Harbin machines (the blue bar/HB in the chart) are going down far more often than the fusion machines at your other facilities.

Driving Improvement

After determining the cause of your underlying issue, you can quickly operationalize improvements by giving broader visibility to this metric. A number of customers have found that they can greatly improve performance simply by giving machine operators and plant managers visibility into real-time key performance indicators.

You click the Save button in the upper-right of the Downtime Count by Day for fusion machines chart to add it to your Operator Dashboard.

The Sight Machine platform allows you to quickly take visualizations and incorporate them into your dashboards. Now each operator can see how their team is doing or how all of the fusion machines at your facility compare to other facilities.

You also want to give your operators a view of how their individual performance compares to others, so you copy this chart by clicking the three dot menu in the upper-right.

Next to Assets, you select the three Harbin fusion machines, as well as the top-performing fusion machine at another facility. After doing that, you can see how the operator’s machine is performing against other machines and this corporate-wide benchmark.

Next to Compare By, you change the view from Compare By Factory to Compare By Machine.

Now you can see how your machine compares to others. You could have set a benchmark target in here as well to give each operator a view into real-time performance against a target.

You click the Save button in the upper-right of the Copy of Downtime Count by Day for fusion machines chart to add it to your Operator Dashboard.

You can also add this widget to the Plant Manager Dashboard, as fusion machine performance is a key short-term metric you want to observe.

You can add a widget, using Sight Machine’s point and click interface, to build a chart off of real-time data so that everyone is working off of the same information. For more information, see Editing Dashboard Widgets in Enterprise Manufacturing Visibility (EMV).

This type of real-time operator and plant manager feedback can increase performance if operators and managers know how they are executing vs. others in real time.

Related Videos

The following EMV product videos are available here:

  • EMV – In Action: Identify, investigate, and drive continuous improvement with Sight Machine’s Enterprise Manufacturing Visibility.
  • EMV – Dashboards Overview: Use the industry’s only out-of-the-box dashboards fed directly by contextualized real-time data for every factory across the enterprise.
  • EMV – Dashboard Builder Demonstration: Build a customized dashboard for factory managers, line operators, process engineers, and the executive team.