Visplore version 1.7.

Featuring a new license type for consumers, new analytics, several extensions to the integration with the AVEVA PI System, and much more – Visplore version 1.7 is an important upgrade you should not miss!

New license type for ‘consuming users’: easy live views on your data

Would several people in your organization benefit from easy access to advanced analytics results from your operational data sources, but don’t require (or shy away from learning) the full capabilities of Visplore?

Visplore 1.7 introduces ‘viewer licenses’ as a new license type in addition to ‘full licenses’. At affordable costs, viewer licenses offer a subset of the full feature set mostly for two use case scenarios:

  1. Consuming custom analyses on live data sources:
    While your power users may author custom analysis templates using the “story page” feature, consumer users can simply open them at a click, getting easy access to KPIs, powerful charts, results from advanced pattern extraction, and much more – essentially without training effort. In contrast to ‘Visplore Free’, the new viewer license type also supports accessing the most recent data from live data sources when opening such analyses, so that everyone in your organization stays up-to-date.
  2. Easy ad-hoc time series viewing: The ‘viewer’ license also provides a simplified cockpit that’s intended for easy ad-hoc inspection of time series from a user-chosen data source. With minimal complexity, users may analyze live data as well as large historic datasets. Zoom and pan at high speed, flexibly filter out time periods, compare time periods, and much more.

Viewer licenses are available in various package sizes and come as personal licenses (no shared seats). Switching from a viewer license to a full license does not require a re-installation, just entering a different license code. Viewer licenses require that Visplore uses our cloud-based licensing system.

Robust extraction of events for noisy data

Defining conditions on noisy sensor data can be tricky – even for simple thresholding, let alone more complex things such as model-based anomaly detection. For robustness, previous Visplore versions already introduced features to consider events only if they exceed a user-definable minimal duration, and to ignore short gaps. Version 1.7 makes it even more intuitive to get robust results: A new optional ‘time window’ for the minimal duration lets you easily constrain your condition. For example, you can express “count as violation, if temperature exceeds 100°C for more than 2h within a 6h window.” This will capture situations when the temperature exceeds the threshold too often, without requiring that it exceeds it for more than 2h in one go any time.

Detecting steady process periods

When analyzing data from a technical process or asset, you often need to distinguish between steady process periods and transients such as jumps or strong oscillations. In modelling, for example, it is common to consider only the stable process periods. The new function “SteadyPeriod” in the formula editor makes it very easy to detect periods when variables remained within a range for a user-specified period of time. This may be used to focus the analysis on these periods, or – alternatively – to detect and compare periods which are non-steady.

AVEVA PI: Retrieve time-weighted averages, totals and other aggregates per time period

When retrieving tags from a PI system, Visplore samples the data in a specified time resolution, such as 1 value every minute. For tags representing quantities such as energy consumption or accumulated production volume, however, you typically require the data to be meaningful aggregates rather than samples. For example, you might want to retrieve energy consumption as time-weighted totals per hour. Or bands of the hourly minimum and maximum value. Version 1.7 now lets you specify for each tag, how you want to retrieve the data – sampled, or using one of several aggregate functions. This opens up many new applications in energy reporting and beyond.

AVEVA PI asset analytics made even easier

Using the PI Asset Framework, Visplore makes it easy to compare assets regarding their performance, to search for anomalies across assets, and much more. Version 1.7 significantly improves the handling of the built-in asset browser. It is now much easier to change the set of analyzed assets and tags, and a new filter makes it easier to search for specific attributes.

Computing the time until next event for easier root-cause analysis

Conditions in Visplore now offer to express the time until the occurrence of the next event as computed variable. In case the condition refers to anomalies such as machine stand-stills or process interruptions, this variable representing the time-to-failure can greatly facilitate the root-cause analysis. For example, it can be correlated with all process parameters to spot potential influencing factors. In the image, the blue signal’s jump upward correlates with the decreasing time-to-failure signal (orange), indicating a possible relevance for explaining the impending red oscillation event. As another approach, users can easily compare the last hours before the anomalies to some reference periods when everything was performing normally.

Additional views for comparing KPIs of patterns

The extraction of patterns such as batches and ramp-ups, as well as the computation of KPIs from these patterns, are key features of Visplore. Version 1.7 adds additional views to the cockpit “Pattern Search and Comparison”. Without switching to another cockpit, you can now visualize the distribution of KPIs extracted from the patterns as histogram and box plot, and may use these views for comparing assets, campaigns, and so on by the KPIs.

Various usability feature improvements

  • Changing inputs of computed data at any time: Version 1.7 enables you to change inputs to computed variables any time. So if you discover that you need additional inputs for your formula, just add them, or replace existing ones. This was frequently requested by users and we listened closely to your needs.
  • Right click access to variables: A right click on variable names now offers a convenient way to rename a variable, edit an underlying formula (for computed variables), and to copy the variable name to the clipboard. For variables coming from PI, it also offers the ability to copy the original tag name or path in PI, for a smooth workflow integration with PI tools. This works in most views listing all variables, such as “Statistics” in the cockpit “Trends and Distributions”.
  • Building regression models in every cockpit: It is now possible to build multivariate regression models in all cockpits. This speeds up and simplifies workflows when you need to have model dependencies, for example for anomaly detection.
  • Accelerated pattern search: The performance and robustness of the pattern search feature were improved significantly. This also includes the possibility to define patterns based upon multiple intervals, which was introduced with the previous version 1.6.