Connect machines to the Cloud to reduce maintenance costs thanks to our Machine Learning algorithm.

Approaches to maintenance

Machinery maintenance has always been a critical aspect in the management of company resources. Here it is essential to choose the best approach between Reactive, Preventive or Predictive Maintenance that is able to reconcile costs and number of interventions.


A first traditional approach is the Reactive Maintenance, i.e. going to repair the damage as soon as it occurs. The weaknesses of this model are clear, as, as shown by the graph, the unpredictability over time and the intensity of the damage make it the most expensive solution in terms of resource management and related costs.


A first solution to the problem of unpredictability is obtained with the Preventive Maintenance approach, in which checks are carried out and measures are applied at set time intervals. However, here remains the problem of an unexpected failure that can block production without warning.


With the advent of industry 4.0 and the predominant role of the IoT, a model has developed that manages to overcome the limits of unpredictability of previous approaches. This is the case with Predictive Maintenance. Thanks to the help of connected devices, it is possible to have a continuous, real-time and low-cost monitoring of the condition of the machine. By analyzing the data that are continuously sent by the equipment, specific Machine Learning algorithms are able to understand when the machine is not working well and also to predict a possible failure.

And it is precisely in this last direction that we are moving to

We use engineered and in-house IoT devices that communicate and analyze data from machinery in the Cloud in real time. Thanks to ad hoc trained Machine Learning algorithms, it will be possible to predict any system failures in order to reduce maintenance costs and man hours.

how we achieve it


Provisioning and configuration process of the IoT plant that will interface with the resources to be monitored. The sensors will be adapted to the specific characteristics of the machine for maximum detection accuracy.


Data acquisition in
real time from our IoT Device, which communicates with our Cloud PROCYB® ENGINE IoT infrastructure, allowing for the acquisition of data in real time. It can be supplied already integrated into newly developed hardware, or it can be integrated into third-party hardware.


Real-time control of the status of the machines, managing assets both in terms of resources and components of the IoT plant.

Definition of alarm thresholds following the occurrence of particular conditions dictated by our Machine Learning algorithm, which will predict potential future risks and suggest the measures to be taken.


Once the data has been acquired from the machines, PROCYB® ENGINE IoT processes and aggregates the measurement and monitoring data, to analyze their health conditions. It will then provide the results in dashboards and reports to keep the KPIs of interest under control, so as to extrapolate useful information for predictive maintenance, optimization and cost reduction.


All this is supported by our Machine Learning algorithms capable of applying statistical models, probability calculation and artificial intelligence algorithms to the acquired big data. Machine data, historical data related to faults, repairs, operating conditions, maintenance requirements and additional data sources will be analyzed, allowing production or maintenance managers to predict the failure, identify performance anomalies and perform the root cause analysis.