Product Updates - July 2018

Overview of Changes

Component Type Notes
Action Builder New Feature
The platform is now equipped with the ability to schedule or build actions through integrated analytics plugins to enable the new customizable alerting service.
See Alerting.
FactoryTX Enhancement The FactoryTX configuration file is now automatically backed up to a Google Cloud Services storage instance.
See FactoryTX Configuration Management.
AI Data Pipeline Enhancement In any model configuration in the AI Data Pipeline, you can now view the differences between the Draft and Deployed versions. You can use the Hide Changes and Show Changes button to toggle this option off or on.
See AI Data Pipeline - Show/Hide Changes.
Data Visualization pages Enhancement When using the Compare By feature in Data Visualizations, you can now pivot the data that is displayed in the Table View.Also, the Chart Type menu now includes Grouped Bar and Stacked Bar options.
See Data Visualization - Pivot Table View and Group/Stacked Bar Charts.
Model calendar menu Enhancement When analyzing models, the calendar menu now includes a Time Zone selection.
See Date Range Selection - Time Zones.

New Features

Alerting

The Sight Machine platform is now equipped to integrate analytics plugins with our new customizable alerting service. In order to support this integration, we are also releasing a new version of the analytics infrastructure, which has changed to support a variety of outputs and ensure flexibility for developers.

The alerting service is responsible for polling an analytics plugin on a specified frequency, which can then respond by adding alerts to a running log. In this new architecture, the analytics plugin is responsible for deduplication and attaching metadata that might be needed in user notifications or other customer systems.

The alerting service also monitors this running log, and notifies users of new alerts when appropriate. Currently, the notifications are processed by sending an email to a configured list of recipients. You can configure and customize these alert emails to meet your organization's needs and to show appropriate data related to the analytics plugin.

Currently, only Sight Machine engineering and data teams can build compatible analytics and configure alerts on top of them.

For example, after migrating Sight Machine’s Data Discovery: Timeline Analysis to the new analytics infrastructure, Sight Machine teams could configure alerts for specific machines’ cycle data. The analytics plugin is configured to monitor Diecast machines and send an email alert for every cycle in which Injection Pressure Max exceeds 40psi.

Below is an example alert email.


Enhancements

FactoryTX Configuration Management

When editing the configuration of FactoryTX in the web application or the filesystem, a copy of the configuration file is now automatically backed up to a Google Cloud Services storage instance. This prevents data loss in the case of unexpected failures to the Edge device or FactoryTX container.

The feature does not support automatic deployment of the FactoryTX configuration from cloud to Edge. In case of a failure to connect to cloud storage, the FactoryTX user interface displays an error message, but you can still edit and submit the configuration.

The feature is enabled automatically after you submit a configuration with a valid API Key and API Key ID.

AI Data Pipeline - Show/Hide Changes

When updating an existing model configuration in the AI Data Pipeline, the user interface now displays the differences between the Draft and Deployed versions.

For example, if you add, edit, and delete shifts on a Facility, you see all of the changes highlighted.

If you do not want to see the changes highlighted, on the model’s main page, on the left, click the Hide Changes button. To view the differences again, click the Show Changes button.

Data Visualization - Pivot Table View and Group/Stacked Bar Charts

When using the Compare By feature in Data Visualizations, you can now pivot the data that is displayed in the Table View.

For example, if you want to measure Output by Shift for three machines, the default table view looks like this.

After you click the Pivot button in the upper-right, each Machine occupies a column instead.

You can also use this option for other comparisons, such as Facility or Shift.

In addition, when using the Compare By feature, on the Chart Type menu, the previous option Bar has been replaced with Grouped Bar and Stacked Bar, allowing you to switch easily between these two modes.

Date Range Selection - Time Zones

The Sight Machine platform stores all of your manufacturing data is stored in UTC. Given the time zone and shift schedule of each Facility, as well as your personal preference, we can perform the correct query and properly convert timestamps for display.

KPIs are calculated based on your production schedule, so you do not have to worry about selecting a Time Zone on the KPI Visibility feature, or when analyzing the OEE data model. We have already taken care of those calculations for you.

When you analyze any other models, including Cycles, Parts, Downtimes, or Defects, the calendar menu now includes a Time Zone selection. The first time that you use it, the Time Zone selector defaults to your local time zone (that is, wherever you are presently located). If you have a different preference, you can select a different time zone: either UTC or the time zone of any facility in your Sight Machine environment. If you select one of these, Sight Machine remembers your preference for your next session. Based on your selection, queries and data displays are automatically updated to use the selected time zone.