Zurück zur Hauptseite / Back to main page

Cloud Computing (SS2025)

This seminar applies the fundamentals of distributed systems and cloud computing in practice. Various hardware and software components are combined by groups of students to form a complex system and to enable the practical application of artificial intelligence.

The semester project goal is to develop an edge computing wildlife monitoring solution. Each group of students will develop, build, test, document, and present a complex hardware and software system that automatically detects and identifies specific species from images and videos to monitor and recognize wildlife (cats, dogs, hedgehogs, birds, mice, rats). The semester project includes these tasks:

• development and setup of sensor nodes with single board computers (Raspberry Pi 5) and camera modules (Raspberry Pi Camera Module 8MP v2).
• deployment of the operating system (e.g. Raspberry Pi OS or Ubuntu) and the object detection software (e.g. YOLO [nice source] or Tensorflow)
• collection of a sufficient number of images of at least three animal species of your choice for the training and testing of the object detection model (e.g. Tensorflow, OpenCV, YOLO) with own hardware (in this case you need a supported GPU!) or with a cloud service (e.g. Roboflow or V7)
• development of a backend to manage the sensor nodes and the collected data. The backend can be deployed as docker container(s) on a Raspberry Pi Kubernetes Cluster with k3s (or a similar solution). The backend Infrastructure will be a robust and scalable High Avalability cluster! of Raspberry Pi 3 nodes. The cluster includes a distributed file system (e.g. Ceph with the Ceph Object Gateway S3 API) or a storage service (e.g. MinIO)
• development of a frontend to present e.g. log information, event messages, and a map of events with proof like images and timestamps. The frontend will be deployed as docker container on a Raspberry Pi Kubernetes Cluster with k3s (or a similar solution) too and should be made with a modern framework (e.g. Vue.js or React)
• use of protocol(s) like REST or MQTT for interaction between the components (frontend, backend, services, etc.)
• implementation of a Telegram notification feature with a Telegram Bot.

This course has no written exam! Instead, your individual grade will depend 100% on your work and the results of the semester project.

• On April 9th each team gets one aluminum case (460x340x160 mm) with hardware that should be sufficient to do the semester project. The aluminum case contains:
• 1x Gigabit Ethernet switch with power cable or power supply unit
• 9x RJ45 twisted pair network cable
• 1x Raspberry Pi 5 single-board computer (8 GB RAM)
• 1x Raspberry Pi 27W USB-C Power Supply (for the Raspberry Pi 5)
• 1x Raspberry Pi 4 single-board computer
• 1x Raspberry Pi 15W USB-C Power Supply (for the Raspberry Pi 4)
• 1x Micro-HDMI to HDMI cable (for the Raspberry Pi 4)
• 6x Raspberry Pi 3 single-board computer
• 1x Anker USB power supply (6 or 10 ports) 60W (for the Raspberry Pi 3)
• 6x micro USB power cable (for the Raspberry Pi 3)
• 1x HDMI to HDMI cable (for the Raspberry Pi 3)
• 2x Raspberry Pi Camera Module 8MP v2
• 1x Google Coral USB Accelerator
• 10x Micro SD cards (32 GB each)
• 1x USB 3.0 card reader for SD and MicroSD cards
• 3x SD/MicroSD card adapter
• 60 spacer bolts M2.5

Schedule of the Course

Course Materials

Contact

The best way to reach me is via email: christianbaun@fb2.fra-uas.de