DevOps principles can be extended from the data center to manage workloads on disconnected or intermittently connected edge devices.
The robot is built on a Raspberry Pi running MicroShift, a lightweight Kubernetes distribution, and exposes a simple Flask REST API for motion control.
The system uses a central OpenShift cluster for development and model training, with Skupper for secure communication and ArgoCD for GitOps-based deployment to the robot.
A JupyterLab workbench is used to define and run an Elyra pipeline that trains a YOLOv5 model on the Open Images dataset to recognize fedora hats.
The trained ONNX model is deployed as a scalable and secure REST API endpoint using the model serving feature in OpenShift AI.
A Python Flask application is developed using a web-based IDE (Eclipse Che) with a devfile to manage the workspace and connect to the inference service.
The robot successfully uses its camera and the ML model to detect a red hat, calculate its position, and navigate towards it in real-time.
ArgoCD manages the robot as a remote Kubernetes cluster, enabling automated, Git-driven rollouts of new application and model versions to the edge device.
ROSEN Technology and Research Center GmbH
Lingen (Ems), Germany
17:34Dean Oren & Stefan Belsch
28:19Thomas Tomow
31:38Ignacio Riesgo & Natale Vinto
19:01Christian Koep
29:34Falk-Moritz Schaefer
30:57Teresa Conceicao
24:48Alexander Schwarz
31:02Daniel Oh
.gif?w=240&auto=compress,format)
Jobs that call for the skills explored in this talk.
Hugging Face
Paris, France
Red Hat
Olympia, United States of America
autonomous-teaming
München, Germany
neuland.ai AG
Köln, Germany
