Pilot 1 - FIDIA: Smart Quality in CNC Machining

Problem description

Vibrations (e.g. due to chatter) are well-known issues in the machining and metal cutting sector. They are responsible for poor surface quality in workpieces. The consequences are:

• The need to finish manually the metal surfaces: time consuming & costly in operations,
• High scrap rates: waste of time and resources.

The main problem of conventional solutions is the separation between quality control process and suppression of machine vibrations.

Vibration suppression is usually based on passive custom-made solutions that are effective only at specific frequencies. Broadband active vibrations control is still uncommon.

In addition, product quality is currently assessed only off-line. In this scenario, if deviations are detected, it is not possible to automatically adjust the CNC process parameters to reduce/compensate such errors. Thus, adjustment procedures rely on machine operator skills and it usually takes a long time to identify the adequate corrective solution.

i4Q Solution

The use case will be implemented in FIDIA’s high speed milling machines production site in Forlì, Italy. The proposed i4Q Solutions will combine advanced vibration monitoring methods, with AI-driven prediction of Quality indicators.

The solution will work at different levels:

• First, a novel add-on kit to monitor the behaviour will be developed and integrated. It integrates accelerometer sensors working very close to the source of vibration.
• The signals will feed AI-driven trained algorithms that predict (in-process) the expected quality of the machined workpieces. The AI algorithms will automatically adjust the process parameters in the case some drift/deviation are detected. The algorithms will also identify equipment components degradation.
• Data coming from the inspection of already machined workpieces will refine the AI models and algorithms. The solution will react in advance to deviations that could cause undesired defects.

The solutions will increase the quality, productivity, and efficiency of the process. The number of rejected parts, time devoted to reworking and manual finishing and costs will be reduced considerably.

Expected results

• Assurance of quality of machined product in finishing stage, through optimization of processing parameters: Final surface roughness under 1 micron
• Reduction of Chatter velocity RMS, through in-process suppression of vibrations:  90% reduction
• Reduced Failed Component Identification Times thanks to equipment degradation patterns identification and forecasting: 50%  reduction
• Quality prediction effectiveness: To be exploited the quality prediction needs to be reliable. Its accuracy needs to be higher than 90%