Solution
Camozzi Digital & Mechatronics carried out an interview on a Spark 2100 machining centre equipped with the latest generation sensors, reviewing the available sensors and the signals detected by the service department, and then divided the machine into 8 main groups on which to focus data collection and analysis.
"Following the philosophy behind Mandelli's project, we adopted an approach that would make iPum@predict effective on any type of machine," explains Cristian Locatelli, Camozzi Digital & Mechatronics General Manager. The subdivision of the system into different groups immediately gave good results, which were confirmed by the tests carried out later. Thanks to these tests performed on different
machines involved in different applications, we collected a significant amount of data and experience, which allowed us to refine the algorithms and make the results more and more reliable.
Specifically, Camozzi Digital & Mechatronics approach is based on three pillars:
- the correct finalisation of the infrastructure required to collect the data and make them available
- data collection, filtering and analysis
- the implementation of physical models and algorithms on a statistical basis
Results
Today more than 25 algorithms operate within iPum@ that constantly monitor the progress of machine parameters. Each algorithm refers to 4 or 5 different measurements that are correlated. The data are collected 24 hours a day and processed with deadlines agreed with Mandelli according to real risks and needs.
In this way, an extremely easy to use tool has been created, which does not require any particular IT skills. Customers who purchase
iPum@predict simply need to access the system via a personalised and protected web page, which contains all the information regarding the operating status of the machines in the workshop. Mandelli's technicians periodically analyse the data collected and report any parameters that deviate from the optimal operation of the system. In this way, the customer has the sole task of reading the signal and intervening by modifying the parameters to avoid excessive wear and tear on a specific part of the machine.