PR.I.MA.

Industrial PRototype for predictive MAintenance

CLIENTE:

Progetto di ricerca aziendale

DATA:

2017

SERVIZI:

Realizzazione di un prototipo per la manutenzione predittiva di un motore elettrico

Predictive maintenance is a type of maintenance that is carried out following the identification of one or more parameters measured and extrapolated using appropriate mathematical models in order to identify the time remaining before the failure.

In particular, we talk about “predictive maintenance” when there is a direct relationship between the value of a signal and the residual life of the piece that emits it. This allows to determine the time left before a fault and plan the intervention to be implemented.

On the other side there is the “maintenance on condition” which instead occurs when this relationship does not exist or  cannot be determined and therefore a limit within which to intervene is preset.

PR.I.MA. è il prototipo per la manutenzione predittiva che abbiamo ideato e realizzato allo scopo di monitorare costantemente lo stato di salute del sistema controllato: al primo segno di degradazione del suo funzionamento mette in atto, automaticamente, una serie di azioni correttive che ne preservano l’operatività, evitando il verificarsi del guasto fino all’intervento del manutentore.

Requirement to satisfy

We started from a very specific goal: to implement a predictive maintenance system for a pre-existing controlled system, integrated by: 

  • programmable logic unit (PLC);
  • human machine interface (HMI);
  • actuators and sensors.

In particular, we have paid attention to the maintenance of the motors of the conveyor belts of a material handling system, starting from monitoring their temperature.

Our proposal

To pursue the final aim, we have chones a decentralized control: the parameters of interest are managed by a new subsystem to which the temperature sensor is connected. This solution communicates with the control and supervision subsystem through the exchange of messages.

System functioning 

The system thus implemented, articulates its operation in the following main subsequent processes:

  1. the IoT HUB, using the temperature sensor, makes a reading every 10 seconds, calculates the average of 6 readings taken in a minute and sends it to the Cloud;
  2. the information sent to the Cloud is processed in order to identify any anomalies within the operation of the engine;
  3. based on the results obtained, the Cloud sends feedback to the IoT HUB which, depending on the nature of the information received, imposes a regular, reduced or safety stop operating mode on the PLC.

Funzionamento del sistema simulato in assenza di anomalie

Funzionamento del nastro in assenza di anomalie

Funzionamento del nastro in presenza di anomalie

Stop di sicurezza del nastro

Results

Thanks to the use of communication libraries such as SNAP 7, the system we have created can be easily introduced on pre-existing controlled systems, without requiring substantial changes to the control logic.

Furthermore, during the experimentation it was possible to observe how the constant monitoring of the parameters and the timely identification of the operating drift, together with the implementation of corrective actions perpetrated automatically by the artificial intelligence module, lead to the evident decrease in cases of unexpected failures and a significant increase in system performance.