Carissa Cirelli (Mechanical Engineering) | Digital Fingerprint: Application of Industry 4.0 towards the Manufacturing of Lightweight Textile Composite Parts (Universität Stuttgart Institute of Aircraft Design)

Image

For twelve weeks in the summer of 2022, I interned with the University of Stuttgart Institute of Aircraft Design (IFB) at ARENA2036, a research campus that brings together scientific and industrial enterprises to explore the future of mobility. I worked on an ARENA2036 project, Digital Fingerprint, for which the IFB is a project partner along with corporate partners Daimler and Bosch. The project aims to integrate intelligent data collection, processing, and transfer into the value chain for lightweight textile components to be used in electric vehicles. When I was not working, I was exploring Germany and Europe with the 9-Euro ticket, eating buttered pretzels, making lifelong friends, and developing a fondness for German language and culture.

I was initially interested in the Digital Fingerprint project, because I was hoping to see how Industry 4.0 principles and technologies could be applied to a specific use case. Industry 4.0 refers to the usage of information and communication technology for networking of industrial machines and processes. This is desirable because smart technologies that can communicate wirelessly have the potential to increase automation in industry, requiring less human intervention to analyze and diagnose issues in production. Industry 4.0 can be accomplished by using Internet of Things (IoT), the network of physical objects that are able to wirelessly connect and exchange data over the internet.

The aim of the Digital Fingerprint project is to demonstrate the application of Industry 4.0 and IoT technologies to the production lifecycle of lightweight textile composite parts which have potential applications in lightweight vehicles. As part of the project, a sensor that measures acceleration and temperature is integrated into a composite textile part, specifically, a structural casing for an electric vehicle’s DAC/ADC power electronics module. This sensor collects production-related data on the part as it is being manufactured. The data is then sent to the cloud, where it is processed and analyzed. In the future, the team hopes to use the data to develop a machine-learning algorithm to detect the occurrence of deviations from production specifications. Once these deviations are detected, modifications and fixes can be made quickly, ensuring production quality standards are maintained and waste from defective parts is minimized. Additionally, the integrated sensor can be repurposed after production to collect data on in-service products. These smart products can help customers monitor the health of their product, for example, by alerting the customer if the part needs to be repaired or replaced.

My internship consisted of two sub-projects which were completed in parallel. First, I developed and optimized the manufacturing process for a glass-fiber preform prototype part. Second, I helped program a Raspberry Pi computer to wirelessly transfer production-related data during the manufacturing process to the project’s cloud-based IoT dashboard.

My first objective was to develop a repeatable process for manufacturing the preform, which had a design and specifications defined earlier in the project. The preform is made of four layers of woven glass fiber molded into a rectangular dome shape. The glass fiber layer second from the top has a BMA456 dual acceleration and temperature sensor ribbon woven into the fabric, and the fabric around the part’s mounting holes is mechanically reinforced with a carbon fiber stitching pattern designed to help the part withstand potentially large mechanical stresses induced in this area. The preform manufacturing process involves three subprocesses: cutting the fabrics according to a specified pattern, stitching carbon fiber into the sensor-containing layer of fabric, and draping the fabrics over the rectangular dome mold. During the draping process, the four fabric layers are interspersed with layers of a binder mesh material, heated to melt the binder, and pressed over the rectangular mold shape, producing a stack of fabrics that maintains the mold’s shape after being removed. While the process itself is simple, it was challenging to make a process that was easily repeatable, met desired standards, and didn’t damage the BMA456 sensor. Not damaging the sensor was particularly challenging because the process involves cutting and stitching fabrics with sensors already woven in. I had to develop tools, processes, and machine files that guaranteed the sensor would not be cut in two or stitched through. Other details had to be considered, such as consistent positioning of the fabric stack within the draping mold, the density of binder mesh to use, temperature and duration of the binder melting step, and positioning of the Raspberry Pi module within each machine. By the end of the twelve weeks, we had developed a repeatable manufacturing process, allowing the team to transition to the next stage of the project, which is to start producing multiple preform parts to generate the data needed for the machine-learning algorithm.

The other half of my work was to help develop the data collection software. We used serial peripheral interface (SPI) to transfer the data from the BMA456 sensor to a Raspberry Pi computer, which would then wirelessly transfer the data to a cloud-based platform, Thingsboard, using a Python-based MQTT interface protocol client. I had never used Raspberry Pi or the Linux OS before this point, so I learned quite a bit about Linux command line over the course of the internship. I also got to learn about MQTT, a popular messaging protocol for IoT applications, and how to use it. To enable the software to be run and monitored during the manufacturing process, I wrote a small script that ran the data collection script upon the press of a button and toggled an LED to indicate data collection in process. The software with my modifications will be used to collect data in the next stage of the project.

Finally, a big reason why I wanted to come to Germany was to experience working and living in a different country. As someone who had only ever left the US once before this summer, this was the most enriching part of the internship and I am so glad I did it. There are so many things I am going to miss about Germany: the bakeries on every corner, the public transportation that can get you anywhere, the half-timbered houses, stumbling upon a cafe in the middle of nowhere while wandering through the countryside, partying in Berlin, and of course, all the amazing people I met and befriended. I was really worried about making new friends, but my colleagues at the Institute and others I met outside of work were extremely welcoming. I learned a lot about German culture through them and made some really great friends who I can’t wait to see again.