Want to know more about this project?
Drop us a line!
Italprogetti
On-Edge Predictive System for
Process Optimization Based
on Real-Time Collection of
Process and Environmental Data
Intro
In the current competitive landscape of the tanning industry, where client companies (often major fashion brands) are imposing increasingly stringent requirements in terms of lead times and delivery coordination, the optimization of production time and costs has now become a top priority.
For this reason, Italprogetti (https://italprogetti.it/ – a manufacturer of machinery for the tanning, food processing, paper, and oil & gas industries) has decided to further develop its leather drying systems by leveraging the potential of advanced digitalization. In this way, the company will be able to provide its clients with industrial machines capable of combining process optimization with the high quality standards that have long distinguished the Tuscan Leather District of Santa Croce–Ponte a Egola.
Challenge
The objective of this project is to develop, within an industrial leather drying process, a predictive system capable of estimating the number of minutes required to reach a precise moisture level. Drying time forecasting is not a fixed value based solely on the moisture content of the hides themselves; it is also strongly influenced by a range of contextual factors, such as temperature and humidity, both inside and outside the drying chamber.
The system must therefore be able to collect product and contextual data, as well as production data, in order to automatically initiate the estimation process based on machine tags sourced from the production system.
SOLUTION
Architecture
To address this challenge, a fully on-edge intelligent system has been developed, capable of operating directly at the plant level. Integrated with the existing unified data collection infrastructure (OPC UA), the system gathers real-time process and environmental data, processes them locally, and returns continuously updated predictions of drying time. It consists of:
- An OPC UA connection system that acquires real-time process data
- A communication logic module to trigger the prediction model
- A drying time prediction module for leather, based on real-time collected data
- A supporting database for storing prediction and process metadata
This decentralized architecture enables data processing and insight generation directly at the point of data origin, without requiring a continuous connection to the cloud. In this way, it provides reliable, contextualized, and real-time predictive estimates, even under complex operating conditions.
On-edge logic
Thanks to the on-edge logic:
- Data is processed close to the machine, minimizing latency and network dependency.
- Predictions are dynamically updated based on the progress of the ongoing cycle.
- The system can operate autonomously, detecting the start and end of the process through specific machine tags.
- The generated information is delivered directly to on-machine systems, enabling immediate operator response and greater integration across production stages.
The Results
OPERATIONAL RESULTS
The solution is currently deployed at the Antiba tannery, which uses Italprogetti drying systems. The integration of the predictive system has enabled:
- Greater accuracy in estimating drying times
- Improved synchronization with other production stages
- Reduced waiting times and operational inefficiencies
- Increased operator responsiveness, thanks to precise and real-time updated information
CONCLUSIONS
This solution represents a concrete example of on-edge smart manufacturing, where operational intelligence is distributed directly on the devices at the production line.
It is an approach that combines advanced digitalization, local autonomy, and rapid responsiveness, with the goal of transforming each machine into an intelligent and proactive node of the factory of the future.
