Testing to verify functional requirements on complex assets can require 24/7 monitoring by specialised personnel. Furthermore, for this type of test it is crucial to ensure proper operation of the infrastructure supporting the test as its failure may not only invalidate the test itself but could cause extensive and irreparable damage to the part being tested.
In accordance with the customer’s needs, we developed a test control system capable of real-time monitoring and management of all the sensitive parameters of the thermoregulation system (essential for proper execution of functional tests and preservation of the integrity of the system being tested). Moreover, the use of AI algorithms allowed for the creation of a control system that safeguards the system in the event of a failure: this is achieved by issuing warnings/alarms to specialised personnel who consequently take action only when necessary, thus resolving the problem of continuous functional test supervision.
The thus-structured system is able to interface with the maintenance CMMS, making management of maintenance tasks predictive rather than preventive.

The main goal of the project was to maximise machine reliability to prevent finished product defects.
The project was developed by an interdisciplinary (Ceflatech/Customer) team comprised of experts in product quality, the process and production technology.
Starting with the defect detected on the product, the exact point in the process at which it occurred was located (process backtracking). ). Identification of the point where the defect was generated was followed by process analysis to determine the key variables in play and understand how machine failures/drifts were able to generate the defect (FMEA, FTA and FMECA). After initial screening of production data and experimental data (specifically designed tests were performed), a multi-factor trial was implemented (D. o. E.) to identify the main factors and their cross-effects in order to reduce process variability.

A galvanizing process for the surface treatment of large items is not just highly expensive in terms of energy: it also involves considerable environmental impact on the waste management side.
The illustrated project is divided into several stages that range from the electrical and structural revamping of several tanks to the implementation of AI algorithms that make the process more energy-efficient.
The plant efficiency enhancement project is a scalable pilot project. Initially implemented on the chromic anodic oxidation tank, it can, over time, be exported to the other stages of the process (the entire process consists of about 40 different stages).
Starting with data made available by the customer, detailed analysis was carried out: this highlighted some correlations between the concentrations of the reagents and the periodic replenishments used to regenerate the tank itself. This analysis was followed by identification of a set of technological parameters (regulated but not saved and logged) to be monitored. Subsequently, sensors suitable for data recording and control via AI algorithms on C-Platform were installed. . In parallel, the ‘intelligence’ needed to optimise the process was developed.
At present, system output is able to modulate demineralized water flow rates as a function of actual system usage: it also controls tank reactivation times and the reagent quantities needed to restore target values.