Quality Mining, the interaction of quality, production, and process

Questioning specifications and making them usable

• The testing strategy does not distinguish between critical and non-critical characteristics, which leads to high testing costs.
• Despite passing the incoming goods inspection, fitting problems occur during final assembly because the actual process risks are not taken into account.

Quality Mining solution:

CTQ-based testing strategy instead of blanket dimensional testing

• Quality-critical characteristics (CTQs) and critical characteristics with a direct influence on function and service life are subjected to intensive testing.
• Non-critical dimensions → Only random sampling or proof of process capability by the supplier.

Dynamic adjustment via structured root cause analysis

• If production problems arise, a systematic cause assessment (IS-IS-NOT, 5-Why) is carried out via the integration of complaint and action management into the worker guidance system.
• Failure patterns are consistently fed back into the incoming goods inspection via changes in the control plan and FMEA.

Result:

• 30% less testing effort, as only CTQ characteristics are tested intensively.
• Improved process reliability, as incoming goods inspections are tailored to real risks.
• No uncontrolled changes to test plans, but systematic adjustments.

Problem:

Installation errors occur because seals are inserted incorrectly. Training only helps in the short term, and rework incurs high costs.

Quality Mining solution:

Digital worker guidance with poka-yoke

• The assembly process is supported by a digital worker assistance system (vision-based, RFID) that is seamlessly connected to the CAQ.
• During the assembly process, poka-yoke safety devices prevent components from being inserted incorrectly.
• The CAQ system documents each assembly step via an interface and prevents subsequent process steps from being omitted.
• Serial number tracking using barcodes

Result:

• 80% reduction in assembly errors, as workers are immediately alerted to problems.
• No waste due to incorrect parts, as Poka-Yoke provides security.
• Digital documentation with serial number tracking

Reducing complexity – Automatic test control & digital worker guidance

Problem:

In the manufacture of structural components, workers must document measurements manually, which is time-consuming and prone to errors. In addition, inspection intervals are defined as rigid specifications, leading to unnecessary inspection efforts or unrecognized process instability.

Quality Mining solution:

Digital worker guidance with test equipment integration

• Test assignments appear mobile & context-related with clear instructions.
• Measurements are automatically recorded and documented via MQTT, without manual input.

Adaptive testing strategies for less effort and greater security

• Inspection severity adapts: Stable processes → inspection rate decreases, deviations → inspection rate increases.
• Real-time verification with the CAQ system to detect errors at an early stage.

Result:

• 30% less testing effort, as testing is only intensified when necessary.
• 100% digital documentation, no paper records or manual errors.
• Faster responses to deviations, as actions and complaints are automatically triggered from the inspection dialog.

Problem:

In the production of medical devices, dimensional deviations occur time and again.
Spot checks often detect errors too late or not at all.
The lack of real-time feedback between machine and quality data means that sources of error cannot be eliminated at an early stage.
In the event of an error, the machine is readjusted and the collection boxes produced are subjected to a re-inspection.

Quality Mining solution:

100% inline inspection with direct machine connection

• Optical sensors detect each unit produced inline and check dimensional accuracy, including documentation of a time stamp.
• If a deviation is detected, the faulty part is immediately rejected.
• The CAQ system stores all test data and sends feedback to the machine control system.

Machine and quality data in a control loop

• Analysis of the data reveals whether dimensional deviations are caused by tool wear, material changes, or process parameters.
• If a pattern is recognized, process adjustments are made (e.g., release checks during machine adjustment or interval-dependent tool changes).

Result:

• Zero-defect strategy, as rejects are detected and eliminated at an early stage.
• Reduction of testing times by 50%, as manual intervention is no longer necessary.
• Higher machine availability and more stable processes through direct quality feedback.

Problem:

A machining company changes tools either too early or too late.

• Too early: Tool costs rise unnecessarily, frequent interruptions occur.
• Too late: quality problems and rejects increase.

Quality Mining solution:

Tool change control via machine data & CAQ

• The CAQ system automatically evaluates machine counters and quality data.
• Once a critical number of components has been processed, a scheduled tool change is triggered.
• If SPC data shows that dimensional stability is changing, the change can be made earlier.

Result:

• Extension of tool life by 20% without compromising quality.
• Automated process without manual planning or monitoring.
• Automated shift logs and evaluations are available