Mondragon Innovation | Humanity at work – Detection of Cutting Tool Breakage in Machining from Vibration Signals with Artificial Intelligence Techniques
Mondragon Innovation | Humanity at work
Sector: Industry
Business Case
Milling cutter tooth breakage in long unattended machining processes can affect the quality of the resulting part and not be detected until it is too late to avoid scrapping it. Furthermore, depending on its severity it can even affect the machine condition.
Objectives
Detect cutting tool breakage from vibration signals during the milling process.
Use case
Focused on two phases: (i) An analysis based on the vibration signal to obtain the variables that can best identify the breakage; and (ii) the implementation of a binary machine learning model.
Infrastructure
On Premise Edge
Technology
Machine learning and deep learning
Data
A vibration dataset from laboratory failure tests, from which the relevant harmonics for failure detection are extracted.
Resources
The technical work was led by a team of two people from IDEKO and in terms of infrastructure, the data collection work was carried out in the laboratory and the field tests were carried out using the DANOBATBOX on a real production machine to which DANOBAT has access.
Difficulties and learning
The large number of different cutter configurations (dimensions – number of teeth) makes it complicated to cover the entire tool catalogue. Laboratory tests were extensive enough to cover a large number of tools with different characteristics.
KPIs (business impact and metrics of the model)
Defect detection / labelled defects
Funding
Private
Collaborators, Partners
IDEKO’s data analysis area was involved in the technological leadership of the use case.
