01
Many machines, many views
Quality data, machine states and events are available. To find changes, your team has to check the individual views.
Data Story · prioritised
- Line 02 · Assembly
- This week vs. previous week
Observation
Rework on line 02 has risen from 4 to 14 percent. The increase affects only variant A; variant B stays at 2 percent.
Rework share of variant A per day, percentThis week: 14 percent
Check: Compare the process parameters of variant A at station 3 with the previous week.
Why only variant A?
Manufacturing Insights · AI Copilot
Which machine stands out? The Copilot examines your production data.
More rework, more frequent stops or changes in process paths: the Manufacturing Analytics Copilot examines available Insights data in the background. It prepares findings as Data Stories. Your team checks the evidence, asks follow-up questions and investigates the case on the machine.
- Background analysis
- Review the finding
- Context-Aware Analysis Chat
Manufacturing Insights Add-on
Dieses Modul beantwortet eine konkrete Produktionsfrage auf der gemeinsamen industriellen Datenbasis.
Time lost before the actual decision is made
The data is available. Who checks every machine each day?
Copilot handles the routine search for relevant anomalies and presents prioritized investigative approaches to the team.
02
Which change matters?
One line produces more rework; another stops more often. Your team needs indications of where a closer investigation is worthwhile.
03
Evidence for the next meeting
The machine, time period and comparison values need to be brought together before production and quality teams can discuss a case.
Automated Analysis with Expert Review
From Insights Modules to Prioritized and Verifiable Anomalies
Copilot automates analysis and data preparation. The decision regarding the cause and the appropriate action remains transparently with the team.
- 01
Merge Module Data
Copilot uses the structured plant and analysis context from the activated Manufacturing Insights modules.
- 02
Conduct an AI-assisted investigation
It checks data for relevant anomalies, patterns, and correlations without requiring each analysis to be initiated manually.
- 03
Prioritizing Problems
Anomalies are organized by relevance and presented along with supporting data and visualizations.
- 04
Track
Your team checks the finding on the machine, decides on an action and then compares results under comparable production conditions.
An example: why is rework increasing?
An investigation with the Copilot can follow this sequence: a finding points to the affected line, a Data Story supplies the evidence and a follow-up question narrows down the products. This example illustrates the workflow; it is not a measured customer result.
Does the change affect all products or just one variant?
Rework increases on one line
Quality data is available for several lines. The Copilot examines it in the background and identifies an unusual development. Your team gets a starting point for investigation without first checking every line individually.
Schematic example with fictional values; not a customer result.
Autonomy with Clear Responsibilities
Automate more of the analysis process. Ensure that production decisions remain transparent.
The Copilot handles the search and preparation. Your team checks whether the comparison is meaningful and which cause can be confirmed on the machine. Events occurring together provide an initial clue.
Automated Analysis with Visible Supporting Documents
The Copilot can analyze and prioritize data on its own. The resulting data story shows which observations and contexts led to the conclusions drawn.
Prioritization Instead of Autonomous Production Decisions
The module saves time spent on searching and processing data. The responsible teams retain the authority to confirm the cause and decide on appropriate actions.
A common context across all modules
Copilot, Insights modules, Chat, and Data Stories all operate on the same system, data, and permissions infrastructure.
Autonomy and Control
What the copilot handles on its own—and what remains the team's responsibility
What can we investigate with incomplete data?
Quality data can show that rework is increasing. Without process parameters or part histories, it cannot yet narrow down possible causes. The Copilot does not supply missing measurements. During the pilot, we identify which questions your available data can answer.
How do we measure the value of a pilot?
Compare the effort of your previous search process with working from Data Stories. Record which findings led to an investigation, which cause was confirmed and whether a subsequent action changed rework or downtime.
What exactly does “automatic analysis” mean?
Copilot can analyze the existing, structured data from the Insights modules for relevant anomalies and patterns without requiring individual user input, and prepare the results as prioritized data stories.
Does the copilot make production decisions on its own?
No. He prioritizes the need for investigation and prepares supporting documentation. The technical assessment of causes, approval, and implementation of corrective actions remain the responsibility of Production, Quality, or Maintenance.
What data can the Copilot analyze?
It utilizes the available equipment and data context from the activated Manufacturing Insights modules, such as OEE and production data, notifications, machine statuses, process paths, and part histories.
Why is a chat feature needed on top of that?
Data Stories proactively highlight prioritized anomalies. The chat feature then allows users to ask follow-up questions, narrow down the results further, and access linked analyses directly.
Start with a clear question about priorities
Choose one line and one question, such as increasing rework. During the pilot, check whether findings are relevant and how much time your team needs to reach a case it can investigate.
- 01Focus
Select a module and a prioritization question
Specify which existing analytics data the Copilot should examine and which type of anomaly it should prioritize first.
- 02Basis
Check the Data and Asset Context
Relevant OEE, notification, status, process, or tracking data is checked for completeness and technical relevance.
- 03Pilot
Testing Data Stories and Chat in Everyday Life
The team reviews automatically generated alerts, asks follow-up questions, and verifies the associated supporting documents.
- 04Scaling
Evaluate Time Savings and the Quality of Prioritization
The search effort, the relevance of the results, and the time required to reach a technical decision are compared before any expansion is undertaken.
30-minute product demo
Which production issues should Copilot prioritize for you?
We'll demonstrate live how the Copilot autonomously analyzes existing Insights data, presents relevant findings as data stories, and answers questions via chat.
Product demo featuring a specific prioritization question