Ferdinand Steinbeis Institute and STACKIT implement AIoT project
![STACKIT & Ferdinand-Steinbeis-Institut Partnerschaft](https://www.stackit.de/wp-content/uploads/2023/01/230125_STACKIT_Partner_FSTI_STACKIT_pw.jpg?x34470)
AIoT Lab team finds every leak
AIoT Lab is a research group of the Ferdinand Steinbeis Institute and is an enabler for the validation of AIoT-capable business ideas. AIoT is an acronym for Artificial Intelligence of Things and is set to play a leading role in the industry in the future. As part of the AIoT Labs, industrial use cases, amongst others, are explored. STACKIT and the Ferdinand Steinbeis Institute are collaborating in a cloud-based project on the topic of “Detection of leakages in pneumatic systems”. The primary objective is to locate leaks in hoses and pipes.
Are you interested in our project results?
Did you know?
![Leckagenwarnung](https://www.stackit.de/wp-content/uploads/2023/01/230124_STACKIT_Leckage_Bild_pw-1-720x447.png?x34470)
A leakage or leak is an untight spot in a system that is triggered for example by a hole in a pipe. This causes the undesirable and unnoticed entry and exit of fluids, gases and solids, barely detected by human hearing. Ultrasonic sensors are mainly used currently to detect leakages. These are only moderately successful in finding such leaks despite a great deal of effort.
Possible fields of application that may be affected by leakages:
![Kaffeerösterei](https://www.stackit.de/wp-content/uploads/2023/01/AdobeStock_50784064-1-1920x1280.jpeg?x34470)
Coffee roasting plant
![Autowaschanlage](https://www.stackit.de/wp-content/uploads/2023/01/AdobeStock_154241758-1920x1271.jpeg?x34470)
Car wash
![Maschinenanlage](https://www.stackit.de/wp-content/uploads/2023/01/AdobeStock_256170723-1920x1280.jpeg?x34470)
Production plant
![Abfüllanlage](https://www.stackit.de/wp-content/uploads/2023/01/AdobeStock_293215427-1920x1280.jpeg?x34470)
Bottle filling system
Leakages can be found in many areas of industry. Besides coffee roasting plants, car washes or production plants, bottling plants can also be affected. The fields of application have one thing in common: Not only does the reduction of pressure losses save energy and costs but also contributes towards reducing CO2.
Figures on per annum leakages in Germany:
More than 62,000 compressed air systems
Over 16 billion kWh consumption of all compressed air systems
Up to 30% loss of compressed air of most plants*
Approx. €5.14 billion savings potential of all compressed air systems*
* according to research conducted by the Ferdinand Steinbeis Institute
Simple illustration of a complex problem
The AIoT Lab of the Ferdinand Steinbeis Institute is rising to the challenge of reliably detecting leakages in pneumatic systems. Leakage data is captured, analyzed, classified and subsequently saved in the cloud using an ultra-sound microphone and an AI solution.
To illustrate the problem of leakages simply, the Ferdinand Steinbeis Institute uses a suitcase-style and true-to-scale LEGO® brick construction of a bottling system.
![AIoT-Koffer geöffnet](https://www.stackit.de/wp-content/uploads/2023/01/230120_STACKIT_ENT_Steinbeiss-Koffer_fh-6537-1920x1280.jpg?x34470)
![AIoT-Koffer nahe Ansicht der Anzeige](https://www.stackit.de/wp-content/uploads/2023/01/230120_STACKIT_ENT_Steinbeiss-Koffer_fh-6515-1920x1280.jpg?x34470)
This suitcase serves for training and demonstration purposes at interested companies in the industry. For the first pilot project, TÜV Süd and Bosch are also partners besides STACKIT Mader. Bosch has provided the sensory technology for the leakage detection.
Advantages of the AIoT Labs leakage model
A number of noteworthy advantages of the leakage solution have emerged from the collaboration to-date. The success of the project has especially been shaped by the co-operation between STACKIT and the Ferdinand Steinbeis Institute. As an EU secure, sovereign and especially regional cloud solution, the STACKIT cloud forms a reliable basis for this model of success. This enables GDPR-compliant data processing and storage globally. In addition, the users have many other advantages:
1. Easy integration of the model in existing maintenance measures |
2. Complete digital package, including push notifications on leakage data |
3. Prioritization of relevant and less relevant leakages |
Not only do these benefits increase energy and cost efficiency but equipment downtime is also significantly reduced with status monitoring and predictive maintenance.
This is how leakage detectors function together with STACKIT Cloud
![Diagram of how the AIoT_Leakage solution works](https://www.stackit.de/wp-content/uploads/2023/01/230124_STACKIT_Schaubild_Steinbeis_AIoT_pw-1024x576.png?x34470)
Simplified representation of how the leakage model works
At the outset is the pneumatic system where, by means of an ultra-sound microphone, leakage sounds are recorded and sent in real time to a central receiver, the edge node (Raspberry Pi).
The foundation of the model is formed by artificial intelligence (AI, an intelligent Deep Learning algorithm), modeled according to an audio-based air leakage data set. This AI is used on the edge node. After testing, the AI extracts and classifies the most relevant features from the acquired data and learns continuously. Collected leakage data are transmitted directly to the STACKIT cloud.
In the cloud, the data is extracted and corresponding leaks are detected and classified. Leakage data is currently collected from four different locations worldwide and further optimized continuously by the AI.
Leakage results are visualized by AIoT Lab via the platform Grafana and maintenance work can start. In this way, the pneumatic system can also be predictively maintained.
Are you using a pneumatic system and want to test the solution on site?
A great deal of leakage data has been collected in the pilot project between the Ferdinand Steinbeis Institute and STACKIT. In this way, the AI is trained further. In order to make the predictions about potential leakages more precise, AIoT Lab is looking for companies like you!
Then there are only four steps to analyze your pneumatic system:
- To contact the AIoT Lab team: Dr. Dominik Morar from the Ferdinand Steinbeis Institute is pleased to arrange an appointment with you to participate in the analysis. He will be happy to answer any further questions about the project.
- Selection of the companies: Companies will be selected according to the criteria set by the Ferdinand Steinbeis Institute to participate in the leakage test.
- On-site visit: The AIoT Lab team of experts will visit you with 2–3 people for 90 minutes and collect all the necessary data for the analysis of your pneumatic system. Your data is stored in the EU-secure STACKIT cloud. As part of the on-site analysis, you will also receive insights into the functioning of the Steinbeis solution for leak detection.
- Your final report: You will receive a detailed final report with evaluations as well as recommendations for leakage reduction in your business environment.
You would like to implement another cloud project and have a question about the STACKIT cloud?
Please feel free to get in touch with our colleagues from the sales team.
Background AIoT Lab and Ferdinand Steinbeis Institute
![AIoT-Lab und Bildungscampus](https://www.stackit.de/wp-content/uploads/2023/01/Aiotlabs.png?x34470)
The AIoT Lab of the Ferdinand Steinbeis Institute was founded in 2021 at the education campus in Heilbronn and supports the validation of AIoT capable business ideas.
Smart connected products and solutions are developed in the Lab to demonstrate how practical solutions can be mastered by AIoT in the real world. The goal is to make it easier and faster to implement AIoT projects by reducing the number of complex requirements. The AIoT platform, which was developed for leak detection is intended to collect initial experience in this area. By reusing the models or algorithms, future AIoT projects should be easier and faster to manage. Further platforms are planned. You will find more information about the Ferdinand Steinbeis Institute and the AIoT Lab under: AIoT Lab – Ferdinand-Steinbeis-Institut
References from other industries
![Exxeta](https://www.stackit.de/wp-content/uploads/2020/12/SuccessStory_Exxeta-1024x683.jpg?x34470)
![tsg hoffenheim tor](https://www.stackit.de/wp-content/uploads/2024/09/20240824_hbr_1946-1024x683.jpg?x34470)
![SchwarzMobility](https://www.stackit.de/wp-content/uploads/2020/11/Platzhalter-SchwarzMobility.jpg?x34470)