Skip to main content

Simulation in Real Time for Manufacturing with Zero Defects

Periodic Reporting for period 3 - STREAM-0D (Simulation in Real Time for Manufacturing with Zero Defects)

Reporting period: 2019-04-01 to 2020-09-30

The vision of STREAM-0D is to prevent defect generation and propagation by minimising product variability, exploiting to the maximum level the potential of simulation models, generating real time running models to optimise the production process through smart decision making, based on the combination of input data from the real world and output data from the simulation model. The core of this approach is constituted by the Reduced Order Modelling (ROM) approach, which allows the generation of real time capable models. These models, fed with actual data measured in-line during the process, will generate the control data for the production line machines in order to adjust the process parameters to get a “perfect” product, fitting exactly the target design specifications.
This closed loop between simulation, measurements and production allows us to continuously adjust the production process and quickly customise it for new batches with different design targets, reducing down times related to readjustment of the line for new designs.
STREAM-0D will allow industries:
• To adjust the manufacturing process in real time, introducing smart decisions based on the forecast of the models, thus reducing product variability, saving time and costs and reducing the number of rejected units or discarded material.
• To absorb the effect of component variability, and even reduce the tolerance requirements for suppliers.
• To exploit the full potential of knowledge-based and data-driven simulation models not only in the design process but also in the assembly process.
• To increase the flexibility of manufacturing processes, by setting design targets online, customizing batches of units with different features, and reducing down time related to changing design specifications
The project has run for 48 months. These are the main results:
In WP1 “Requirements and specifications”, the industrial processes were thoroughly described, the critical quality indicators for each product were identified, as well as the critical process parameters. This information was compiled in Application Project Charters.
In WP2, “Real-time simulation modelling”, complex physics-based models were developed and converted into real-time apps. The final result is a model that runs in real time providing a solution for the output variables of interest as a function of the input parameters.
In WP3, “Online data gathering systems and data-driven models”, innovative sensors and instruments were developed to gather online, in real-time, the input parameters for the ROMs. Some of these input parameters are not straightforward to identify or to measure, as they are “hidden” variables that need the development of innovative ways for measuring them in a non-destructive way. Traceability systems were developed and installed in order to generate correlated data sets that were used to extract data-driven models.
Within WP4 “Adaptive control”, optimisation and decision-making algorithms were developed. The former are needed for the calculation in real time of the value of the process parameters that adjust the performance to the target value, based on the output of the real-time simulation models. The latter generate alarms, warnings and recommendations, based on the predictions of the DDMs. Additionally, recalibration algorithms were developed to adjust the values of unknown ROM parameters based on the measurements coming from the line, to increase the model accuracy. Finally, a data management platform (iSCAN) that integrates the info from the modules and sensors, and hosts the recalibration module and the visualisations (iDashboards) was put in place.
Within WP5 “Integration”, all the previous modules and systems were integrated in the pilot production lines and tested, first offline or in open-loop or through the use of virtual lines, before proceeding with the final tests in closed-loop. User interfaces were built for the three lines to allow users to monitor on-line data (historical and real-time), interact with the STREAM-0D modules and see in real-time the alarms and predictions from the DDMs and ROMs.
Within WP6 ”Demonstration”, the performance of the STREAM-0D solution has been assessed through the manufacturing of a number of parts, including the manufacturing of different references, so that not only the performance of the ROMs and DDMs has been validated, but also the time needed to change the line for producing a new product reference. Other indicators have been analysed, such as the variation in productivity, cycle time, end-of-line tests, production costs, waste and rejections that results from the implementation of the STREAM-0D solution.
Within WP7,“Dissemination, exploitation and IPR management”, a detailed plan for the identification, protection and exploitation of results has been completed as well as a deatiled business plan for the further deployment of the STREAM-0D solution after the project. All the necessary tools and activities for communication purposes have been put in place.
Dynamic Data-Driven Application Systems is the linkage of modelling tools with measurement devices for real-time control of applications. This synergistic feedback control loop has been applied in some sectors, but its huge potential for manufacturing remains untapped. The ambition of this project is to show the technical and economic viability of this novel technological solution in an operational industrial environment, through the following advances:
-Adaptive process control based on real-time simulation integrated in the production chain
-Reduced Order Modelling techniques for real-time simulation and real-time process control
-Fast, robust and accurate measurement of complex dimensional and material properties
-Application of data-based models for real time control and adjustment of production processes.
The expected impact of the project was:
Achievement of zero defects in multi-stage production lines
Reduction of production costs by 15%
Increased production flexibility. Higher production rates by 15%
Reduction of waste and scrap by 10%
After deploying the solution in the three pilot lines and analysing the results, it has been assessed that the impact goes beyond our initial expectations. In particular, scrap has been decreased by 20%-60% and rejection rates caused by issues directly addressed by STREAM-0D have been decreased by over 72%. Downtimes employed in adjusting parameters due to products out of specification or preparing the lines to produce new references have been decreased at least by 50%, lowering also the production costs. Production rates have been increased by 10-15%. Variability in the KPIs addressed by the project has been significantly decreased so that the process capability indicator Cpk is over 1,67, ensuring that 98% of the parts are within the product limits. Yearly savings due to STREAM-0D have been estimated in 150000 € per line for one of the end users, 465000 € per line for another end user and over 1.23 M€ for the whole plant in the third case.
The three products are representative of a broad number of industrial products, processes, materials and physical phenomena of the European manufacturing industry. The project has been demonstrated on these three lines, but the exploitation and dissemination plan will make it possible to leverage the effect of the project results, impacting a large number of industries after the project.
Project leaflets
STREAM-0D workflow
The three end user applications of the project
Project logo