Home Blog AI for predictive ma ...

AI for predictive maintenance


The Italian manufacturing industry, today as never before, is looking for solutions to improve the reliability, performance and safety of its production chains.
On the one hand, there is a strong focus on controlling costs and making the most of the investments made, while on the other, there is a strong interest in innovation and Artificial Intelligence as a tool to combat international competition. Within this very challenging landscape, 3rdPlace (a Datrix Group tech company) helps companies (including SMEs) to obtain the maximum return from the assets they already have.

A particularly virtuous example concerns predictive maintenance where, thanks to the possibility of Big Data analysis through Machine Learning, it becomes possible to monitor the conditions of the entire production chain in real time, promptly capturing signals relative to possible equipment failures and detecting anomalies before they turn into costly accidents (downtime).

Unlike traditional suppliers, 3rdPlace does not offer itself as a reseller of pre-packaged hardware or software, but as a partner capable of offering highly customised solutions that can promptly respond to the customer’s needs.

Starting from the assets, data and time series already available and leveraging open-source solutions, we propose a highly innovative path centred on tangible and measurable results.

The case: AI to make production more efficient by reducing downtime

Starting from a production chain already equipped with various sensors located in different areas (such as thermocouples, pressure switch, accelerometer/vibrations, current clamp, cameras, microphones, etc.) and having defined with the large client company the KPIs to work on to improve production, the first step was to channel all the data into a single container (datalake) to make the data crossable with each other.
Then, by leveraging the collected data, it was possible to train different machine learning algorithms to monitor and improve the KPIs of interest.
Finally, a web dashboard was delivered to the client, capable of describing the status of the entire supply chain and highlighting any anomalies and criticalities.


In just 6 months, we started with the collection of requirements and then completed the data modelling study, bringing into production an AI-based solution that made a difference. One year on, the results are very tangible:

  • 3 million Euros saved thanks to “early-warning” on anomalies and material damage,
  • 30% reduction in maintenance costs,
  • 25% increase in production,
  • 25% reduction in downtime./li>