29^th November 2021¶

Scope¶

We have worked hard to deliver a great MindSphere Private Premium experience, but we are still tracking some known issues. If you find others, please give us your feedback by contacting our MindSphere Support.

We recommend that you read the following release notes carefully. These notes contain important information for installation and use.

These release notes are relevant for MindSphere Private Premium Release in November 2021.

MindSphere Private Premium¶

Within this release of MindSphere Private Cloud, Semantic Data Interconnect, Digital Twin and Closed loop framework, Operations Insight Analytics and Predictive Learning are included.

These functionalities are part of the premium package offering of MindSphere Private Cloud.

Semantic Data Interconnect¶

Siemens Semantic Data Interconnect(SDI) allows organizations to maximize the value of enterprise data from disparate sources along with IoT data.

The SDI framework provides customers the ability to correlate data from disparate systems and gain powerful insights which would not be possible by having the information in disjoint applications and their storage.

SDI does this by providing the following capabilities:

• Automatic schema discovery

  • SDI follows Schema-on-read philosophy
  • SDI extracts metadata like data columns, data types, etc. without human intervention
  • Custom Data Types: Ability to define custom data type fields for example: IDs which can be discovered and tagged to corresponding data points
  • Reduces time and effort for system experts to understand various enterprise

• User friendly native SQL Interface

  • Build and consume custom KPI calculations for custom needs
  • Faster learning curve for data analysts to develop queries based on use-cases
  • Enhanced operators to support wide variety of use-cases

• End to End Data Management

  • Well defined Data Management (DM) strategies for data lineage
  • Part of data source registration process and currently supports two DM policy:
  • Append: Data will be appended/added with every data ingest or data generated for source-data tag or Asset-Aspect combination.
  • Replace: Data will be completed replaced with every data ingest or data generated for source-data tag or Asset-Aspect combination.

Digital Twin and Closed Loop Framework¶

The Closed Loop Framework and Applications enable users to realize Digital Twin using the models from the Enterprise application and MindSphere IoT Data. The products are:

• Foundation

  • Manage enterprise model definitions using JSON based schema or industry standard FMI 2.0
  • Manage connections between virtual models and onboarded assets
  • Configure connectors to connect enterprise applications via secure communication
  • Enables custom connector development

• Product Definition

  • Connect specific product configurations to the onboarded assets
  • View the summary of operational events generated for all the product configurations
  • View operational events and related product defects for a specific configuration
  • Create new product defects for the unaddressed events
  • Access to PLM application Interface with PLM applications by using the available or custom connectors

• Discrete Events Simulation

  • Connect discrete events simulation models to the onboarded assets.
  • Manage simulation model parameter configurations for operational data requests.
  • Request operational data from onboarded assets for any time period and configuration
  • Execute Simulations in interactive and non-interactive mode.
  • Execute simulations using operational data for playback and analysis.
  • Interface with any discrete events simulation application using available or custom

• Visualization

  • Connect system simulation models to onboarded assets.
  • Manage simulation model parameters within the application.
  • Create multiple simulation configurations to run battery of simulation tests using operational data simultaneously.
  • Configure variables whose simulation results are stored in IDL.
  • Visualize simulation data within the application.

Operations Insight Analytics¶

Operations Insight Analytics allows you to Define and analyze operation performance through KPIs and drill down by different parameters (machine, product, scheduling, process etc..) and bring manufacturing data together with performance data and other external data to provide instant traceability of field performance issues and tie back to manufacturing or design. It also provides insights that relate part issues to machine failures, machine failure to product quality and manufacturing process to high failure rates. This is realized using the following capabilities:

• Simple KPIs

  • Simple KPIs provide a quick means of monitoring machine or production line data by running the KPI on a simple schedule and writing the KPI calculation results data to a user-selected destination.
  • Simple KPIs are written in JavaScript, can include conditional loop statements, and be scheduled to run automatically on the day and time you select.
  • Simple KPIs are processed internally by Visual Flow Creator and utilize an output variable that must be created beforehand as part of the asset model created in MindSphere Asset Manager.
  • Configuration options include:
  • Up to five input variables, but only one output variable
  • Simple or advanced calculation cycles

• Advanced KPIs

  • Advanced KPIs utilize a KPI format with easy-to-use features, such as drag and drop functionality for KPI components, and operators that cycle through +, -, /, * by clicking.
  • Advanced KPIs support calculation filters to include or exclude specific values or ranges of values from the source data.
  • Advanced KPIs currently supports aggregated time series and non-time series data stored in IoT databases only.

• Advanced KPI Analytics

  • Adding an analysis to a KPI provides additional configuration options, including:
  • Sorting and filtering results table data individually by column
  • Customizing how the analysis results are grouped on the chart's X-Axis (and table columns)
  • Drilling down into analysis data
  • Viewing the history of the KPI's field combinations
  • Scheduling the KPI to run automatically
  • Setting a value threshold that, when crossed, notifies you
  • Setting up an auto-discovery configuration to run when a threshold is crossed

Predictive Learning¶

Predictive Learning allows you to build predictive models through machine learning techniques, enabling companies to optimize product quality as well as reduce potential field failures and performance issues. The features include:

• Simplified model execution with Job Manager

  • Enhance analytical insights from model execution from both PL and imported models.
  • Select required MindSphere data set, model, and compute environment to execute each job
  • Improved visualization and reporting from model execution
  • Workbench is not a part of the private cloud offering.

• Import various types of data sets

  • Increased control for data scientists to import and search various data sets
  • Refined search capabilities to define assets, aspects, and timeframe
  • Import IoT time series data in bulk from various sources
  • Enhance analytic insights into IoT data with Product Intelligence data

• Jupyter Notebook

  • Enable data scientists to build complex or simple models as needed
  • Flexible tool that supports commonly used coding languages
  • Jupyter Notebooks create an interactive computing environment for writing and running code

• Out-of-the-box functions to accelerate data transformation

  • Perform advanced analytics quickly without the need for additional coding
  • Increase consistency and repeatability of analytics functions in a shorter amount of time
  • Pre-configured Exploration, Analysis, and Transformation panels for data sets.

Except where otherwise noted, content on this site is licensed under the Development License Agreement.