Manufacturing Insights
Your machines generate data. Manufacturing Insights helps you use that data to identify specific ways to improve your production: less downtime, less rework, and more usable capacity. Ready-to-use modules for production, quality, and maintenance are available for this purpose. You don’t have to build the reports into dashboards yourself.
The modules utilize the data connected to and processed by the IoT Data Hub. This allows you to further investigate an unusual production metric: at which station the problem occurred, which products are affected, and what was happening on the production line at the same time.
The Right Analysis for Your Question
Not every analysis starts with the same metric. Choose the starting point that best fits your current problem.
| You want to know… | This is what you’ll be working with |
|---|---|
| Why are there fewer good parts or more rework? | Production analysis links volume, quality status, cycle times, and process parameters. |
| Which issues should we address first? | The event analysis shows the frequency, duration, and timeline of machine messages. |
| Where are we losing production time? | The state analysis shows when and for how long machines are producing, undergoing maintenance, being set up, or experiencing malfunctions. |
| Which parts go through additional steps or rework loops? | Process analysis reconstructs actual workflows and highlights different process variations. |
| Is our biggest loss in availability, performance, or quality? | OEE brings these three factors together and provides a starting point for a detailed analysis. |
| Which system needs our attention right now? | The Manufacturing Analytics Copilot analyzes data in the background and presents relevant anomalies as data stories. |
From a Production Problem to a Concrete Starting Point
The strength of the modules lies in their interaction. A low output can result from a shutdown, slower processing, or additional rework. The total quantity reflects the result. Further analyses help narrow down the cause.
Here's an example: A production line is no longer meeting its usual output target.
- Check quantities and quality. In the production analysis, you can see whether fewer parts are being processed overall or whether the percentage of scrap and rework is increasing.
- Identify lost time. The state analysis shows whether individual stations require more frequent maintenance or experience longer periods of downtime.
- Investigate malfunctions. For periods showing anomalies, review the corresponding messages. This marks the start of the technical investigation into specific events.
- Track affected parts. If additional processing is suspected, the process analysis shows which parts pass through stations multiple times.
- Check the results. After implementing a measure, compare the affected quantities, times, or processes again.
You don't have to go through every module. If the issue has already been clearly narrowed down, you can proceed directly to the appropriate analysis.
A shared database for all modules
Manufacturing Insights uses the IoT Data Hub's plant model. Data sources and analysis configurations are assigned to the respective plant. Three technical protocols form the basis:
| Database | What it describes | What it is used for |
|---|---|---|
| Production Log | Parts, processing steps, quality status, and available process values | Output, quality, cycle times, and part histories |
| Machine State Log | Statuses and Their Chronological Sequence | Productive Time, Wait Time, Setup, and Downtime |
| Event Log | Errors, warnings, and other events with system and time references | Recurring faults, event duration, and event sequences |
Existing product characteristics—such as the part type or the tool used—supplement the evaluation. This allows for a targeted examination of differences that would otherwise be lost in an overall value. The specific detailed analyses available depend on the activated modules and the data that is actually provided.
This mapping serves a practical purpose: it assigns a clear meaning to a technical condition code, a readable text to a message ID, and a reference to the correct part for a production event. These basic elements are set up once for the plant and are then used in the analyses.
Start with a specific loss
A good place to start is a production line where your team is already aware of a problem. For example, a workstation that experiences frequent downtime or a product with a high rate of rework. Using this example, you can quickly determine whether the data accurately reflects actual operations and which analyses will be most helpful.
To this end, work with Production and IT to clarify the following:
- Which facility or station are we looking at?
- What aspect—quantity, time, or quality—do we want to improve?
- What data is already available for this?
- Which well-known event can we trace in the analysis?
Bytefabrik supports you with integration, data preparation, and module setup. Upon request, we can also assist with the subsequent analysis and further development of your application. This ensures your team isn’t left to deal with an unusual metric on its own.
Understanding Improvements in Everyday Life
For each measure, document what was changed and what results you expect. Then compare the same production line and, if possible, similar products using the same metric definition. Fewer rework parts, shorter downtime, or a more stable cycle make the progress tangible.
In a data story, you can document the observation, related charts, and the solution all in one place. This facilitates coordination between production, quality, and maintenance, and helps when a similar issue arises in the future.