Projects:

An industrial production line was modeled on a software platform using the industry-standard Siemens Tecnomatix Plant Simulation software. Electrical transformer manufacturing was selected as the industrial case study. The production line system was modeled as a digital twin using actual data such as operation time per part and machine failure rates.

Using the digital twin, multiple scenarios considering low and high demand for transformer production were simulated, and production line parameters were analyzed. Bottlenecks and other machine inefficiencies were identified by running multiple simulations using the digital twin.

An algorithm using multiple fuzzy logic controllers was then implemented to monitor the performance of selected machines. These controllers were responsible for selecting and stopping the appropriate machine combinations. [Serving as a decision support system]

The suggestions provided by the proposed algorithm were able to increase the overall throughput of the production line during the observed 20-day simulation period. Different case studies were used to ensure the robustness of the algorithm, including combinations of random machines, fixed operation of best-performing machines, and fixed performance of average-performing machines. The output of the digital twin was validated statistically to ensure that it closely reflected the actual production line outputs.

Usability of the proposed system:

The connected digital twin will assist the production manager to work in his office rather than visiting the site to make intelligent decisions about machine operation and control, which enables remote production monitoring, which is an important feature of digital twin.

The image below shows the digital model of the transformer production line with consideration for human ergonomics.

The demonstration of the digital twin of the production line is shown below. The production line was modeled using Siemens Tecnomatix Plant Simulation Software.

 Project demonstration, presentation, and result validation: