Home >Open Calls >Experiments > ProcessHealthCheck
ProcessHealthCheck

EUH4D Federation for the development of a service for Data-driven Process health check for industry

COMPANY
Nissatech
DOMAIN
Manufacturing
COUNTRY
Serbia
YEAR
2023

DESCRIPTION

Process Health Check is an AI-based system which can be used for analyzing the performance of a manufacturing process and defining the improvement potential, supporting “the competitiveness and performance of the EU’s industry” as defined in Manufacturing Data Space Challenge (Guidelines for Applicants).

A very important outcome of this system is process behaviour understanding, leading to a better validation of the process improvement potential, a critical activity for manufacturing SMEs. In a manufacturing data space, such a system can be used for a fair comparison of process performances of different systems, leading to standardized benchmarks in Data Space, one of very important functionalities for creating a trustworthy environment.

More generally, this system enables the SME transformation toward “zero defect zero waste” production, which is still missing for most manufacturing SMEs. It appears that manufacturing SMEs have been very slow in implementing formal quality models and environment monitoring tools, due to the need for expert modelling and analysis (expensive and time consuming). Moreover, the problem is challenging due to a huge amount of quality and environment-related data that should be analysed in the real-time, which is unsuitable for traditional process monitoring.

Main Objectives

The vision of this proposal is to develop a service for performing process behaviour analysis based on past process data which can be used as a self-assessment test by manufacturing SMEs in their transformation in data-driven economy.

Main innovations

Main conceptual innovation is related to the modelling part, i.e. how the data analytics methods are used for learning models that can explain what is happening with a system in order to understand how the system will behave in some quality-control related situations. The main issue is the complexity of the models since the process space can be very huge (100+ parameters) and the models should enable very fast interpretations of the data/streams coming from the selected industrial system.

Impacts

The implementation of the experiment contributes to the creation of employment and new job profiles related to the Data sector
By applying services like ProcessHealthCheck, it will be possible to understand that the behaviour of a process was instable in a period and all data generated in that period should be analysed in a different way. This will provide a new quality in the industry data analytics.

Key Exploitable Results (KERS)
  • Process Health Check system that creates process health/behavioural model used for a) data-driven self-assessment of process performances and b)- creation of run-time process monitoring and controlling services (like anomaly detection, predictive maintenance).
  • ProcessEcoQualityCheck - service for analysing the relation between the product quality and environmental/energy aspects.
  • A service that can be used for a comparison of process performances, offered as a standardized benchmark in data space.
  • A new approach for validating the potential for data-driven innovations in manufacturing SMEs, focusing on understanding process behaviour from past data and defining process improvements strategies correspondingly.
Technical KPIS
  • Completeness of the realization of ProcessHealthCheck services (100%).
  • Usability and Business Validation (User satisfaction: greater that 90%, UX: min 90%, Business: min 90%).
  • Performance Validation (Improvement for at least 200%).

COMPANY INFO

Nissatech is an innovation-driven SME with strong international cooperation and vision to become one of European top innovators in the domain of advanced AI and cognitive industrial solutions. The main objective is to develop own technological building blocks through an efficient implementation of the cutting-edge research and their usage for resolving very challenging real-world problems in different industrial domains.