Retrieve Specific Data

Question and objective

What events or metrics are available for a specific machine during a given time period? A limited query helps to verify values and narrow down hypotheses before conducting a comprehensive analysis.

Implementation in Bytefabrik

Queries form the basis for tables, charts, and further analyses of stored datasets. The filters and calculations you set depend on the interface you’re using and the data schema. For custom analyses, AI Notebooks provide additional access to platform data.

This page describes the technical query logic. It does not require a standalone SQL editor or any specific SQL interface.

Best Practices

1. Set a specific start and end time.
2. Select the machine in question and the required fields.
3. First, check the individual rows and the amount of data.
4. Only then should you add grouping and aggregation.
5. Record the time period, filters, and calculation definition along with the result.

Example: To compare cycle times, first filter by product group. In addition to the mean, compare the count and distribution so that a few outliers do not skew the overall interpretation.

What You Should Keep in Mind

Event time and storage time address different questions. For shift-to-shift and day-to-day comparisons, the time zone and boundaries must be identical. Limit large queries early on by specifying the time range and required fields.

An empty result does not necessarily mean there is no production data. Check your filters, permissions, storage, and retention period. Continue with charts or AI notebooks.