We provide EU funding and world-class technical support to engineering and technology businesses.

Find out more about who we are and what we do


Stay in Touch

EventBrite logo Twitter logo RSS logo Linked In  logo Email us

Join our mailing list

Join our mailing list to receive notification of CPSE Labs events and upcoming calls

Smart Anything Everywhere

Cyber-Physical Systems Engineering Labs is part of the Smart Anything Everywhere initiative.

This project has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 644400.

DyPro (Real-time big data driven proactive manufacturing based on CPPS paradigm)

Problem and solution

Although production processes are modelled using a lot of domain knowledge, the behaviour of processes during the execution (run-time) remains “unpredictable”. Indeed, there are many factors that cannot be modelled properly in advance, for instance the environmental context, or the variation in the supplied material. Consequently, for any process that is not properly monitored during run time there is a potential danger that it becomes instable or even breaks down. Especially challenging is the monitoring of adaptive processes, which serve as a basis for realizing flexible production. However, real-time process monitoring requires the knowledge about the anomalous process behaviour to enable a proper reaction and keep the process in the stable state. Since this knowledge cannot be easily obtained and is also of a very dynamic nature, there is a need for exploiting data-driven approaches for generating models of both the normal and anomalous behaviour, which can be applied for the real-time process monitoring.

The DyPro approach is a novel concept for enabling dynamic real-time process monitoring that enables transforming processes into very reactive ones, which can sense and react on changes autonomously, even in the case of monitoring adaptive processes. The approach is data-driven, benefiting from an extensive usage of CPS in the production context, enabled through the expansion of Industry4.0.

The main goal of the DyPro experiment was to understand which challenges need to be resolved in order to realize a very efficient and scalable concept of data-driven dynamic process monitoring. The central outcome is the DyPro Platform, which represents a new generation of the big data analytics solutions and which exploits the combination of real-time and batch processing for enabling continuous real-time monitoring of a production process. The key advantage of the solution is the possibility to find anomalous situations (problems), which even the process owner may not know that they exist in the related process (detecting so called unknown unknowns). The platform is based on the D2Lab (Data Diagnostic Laboratory), a data analytics platform for big data processing, offered as a solution as a service tailored for the Industry 4.0, affordable and easy to use. In addition, the experiment proved that there are many opportunities to exploit this kind of monitoring, in various process contexts.

How did CPSE Labs Help?

The developed platform was evaluated in a scenario-based testing approach utilizing the fortiss future factory demonstrator. For this, the fortiss Design Centre had to expand the demonstrator by adding data acquisition functionalities to the legacy control system used in the fortiss future factory. This expansion was not only a key step for the evaluation of the DyPro platform but also extended the use cases of the fortiss future factory. For this experiment, it could be shown that the DyPro platform behaves as expected and detects anomalies with no false positives and no false negatives. Furthermore, also the performance and scalability of the platform showed its suitability for industrial use cases.

Impact

The developed DyPro platform enables a better process understanding, which consequently opens new opportunities for the data-driven process improvement. In particular, new opportunities exist in a market where manufacturers desire customizable and cost-effective solutions, full exploitation of the data for a complete understanding of the whole production process, and easy re-adaptation of the process to customer requirements (especially towards achieving real-time situational awareness and process adaptation). The main business benefit is the improvement in the process quality control by enabling a more efficient process of real-time anomaly detection. Hence, the DyPro experiment enabled Nissatech to extend and validate their technology portfolio and subsequently offer improved product and services to their customers.

Outputs

The DyPro Platform for big data analytics in the domain of Industry 4.0 has been developed. It allows finding anomalies in production processes without explicitly specifying search criteria or defining undesired situations. In order to achieve these results, models that relate sensors and their values to the process context where developed and incorporated into a complex event processing architecture handling the incoming sensor data stream and analysing it.

Design centre

This experiment is supported by our Germany South design centre

Germany South design centre

Technology platforms

4DIAC Logo

Partners

Dates

1st Jun 2016 - 31st May 2017
Funded under: CPSE Labs Call 2