Kode launches SpectralizeR, the bridge between spectroscopy and production

Author:
Kode s.r.l.
Date:
04.06.2024
Topic:
News

Thanks to SpectralizeR, production process analysis can now integrate real-time quality control at every stage of the process.



Kode, creators of Machine Learning and AI-based products and solutions for over 10 years, launches SpectralizeR, a new tool dedicated to the automation and optimization for the whole manufacturing field. This software integrates spectral data analysis with production process data, allowing faster and more precise data analysis, quality control, root-cause analysis and decision making supporting production processes optimization. With SpectralizeR, real-time quality control integrates production process controll.

Spectral analysis is a method that exploits certain physical properties of matter and, in particular, its interaction with specific radiations (e.g., near-infrared radiations in the case of NIR spectroscopy). This interaction provides information about the chemical composition of the sample within seconds. This approach enables the rapid analysis of a wide range of samples without the need for sample modification or the use of reagents, thus avoiding potential damage to the samples.

Spectroscopy in industrial context.

The advantages of this type of analysis are numerous and can be employed in the industrial field to conduct analyses aimed at monitoring, among other things, the composition (and thus the quality) of raw materials introduced at the outset of the production process. Additionally, this approach can be utilized to verify the efficacy of pivotal steps in the production of semi-finished products, as well as to oversee the impact of specific reactions and to forestall or prevent production errors.

All of this is possible and can be transformed into real-time production control (for corrective action, optimization, or even to avoid unexpected downtime) only if integrated with production data analysis and analysis for MSCP (Multivariate Statistical Process Control). This allows for the understanding of the characteristics of the semi finished product and the conditions and causes that generated the situation highlighted by the sensors.

All these advantages of spectroscopic analysis are often limited by closed solutions with probes and related proprietary software, tied to only one type of radiation or conditioned by the integration of probes with software exclusively from the same manufacturer, thus not allowing new data acquired from sensors to be integrated with machine operations readout dashboards or corporate ERPs. Cross-referencing the data and making a joint reading thus becomes impossible.

Manual quality control

SpectralizeR: real-time quality control integrates process control.

SpectralizeR, to ensure real-time quality control integrated with production process controll, on the contrary, can connect with probes of different types (based on spectral data derived from different techniques, including NIR, XRF, Raman, Fluorescence) or developed by different producers, guaranteeing as such compatibility and freedom of choice.

This software is also capable of connecting with probes from different manufacturers, ensuring compatibility and freedom of choice. Likewise, it easily connects with the various data sources of management systems (OPC-UA, DCS, SQL/NoSQL databases, API, remote folders in FTP/SFTP), extracting key information to obtain a complete and detailed view of the process.

The computational core, based on R, performs spectral analysis, data preprocessing, Applicability Domain and Calibration modeling. Furthermore, it integrates spectral information with process data to implement root cause analysis models and decision support, optimizing the process and correcting any production drift. The analyzes are accessible via a web app which, through a modern and user-friendly graphic interface, allows the selection and creation of models, displaying the spectra and estimated values for the last analyzed sample in real time.

The system architecture is based on Docker microservices, using the Princess proprietary framework by Kode, for access control and security and access to data sources. Its flexible structure allows for its use in both the production environment and remotely, thereby conferring significant added value upon both manufacturing companies and their operational personnel, as well as upon manufacturers of analytical instruments.

Marco Calderisi, CEO of Kode, comments: “Unlike other softwares that focus on a single side of the analysis (such as spectroscopic analysis of production efficiency monitoring), SpectralizeR was created with the aim of creating a point of contact between spectroscopic analyses applied in industrial contexts and the production process itself. The fusion of these two systems offers a complete vision of production, making it possible to overcome the limitations of segregated approaches, which do not fully exploit the reciprocal potential of a unified database”.

SpectralizeR is currently in use by a company that operates in the agronomic sector to monitor in real time the moisture content of different types of products at multiple stages of the production process.

Learn more about our product

Contact form

Thank you for your message