As participants in this interactive workshop, you will develop a robot control application that recognizes red fedora hats using a camera and an ML model. You will learn how to use OpenShift, OpenShift AI, GitOps, Pipelines, Service Interconnect, DevSpaces, and MicroShift in a playful way.
Participants will follow a guide that will introduce them step by step to the challenge. First, they assume the role of a data scientist to train the object detection ML model. Then, they switch to the role of a developer and use the ML service to have the robot actively search for Fedora. During this development phase, the robots will be accessed directly via a Service Interconnect tunnel connection from OpenShift. Using CI/CD and GitOps, these services will be built as ARM architecture-based images and deployed to the robots' MicroShift environment.
At the end of the workshop, participants will have built a complete edge deployment according to MLOps principles, from development on the central OpenShift cluster to edge deployment on the robots, which will then search for Fedoras fully autonomously and compete against each other.
- Rendered Lab Guide : https://cloud-native-robotz-hackathon.github.io/
- Facilitator Guide : https://github.com/cloud-native-robotz-hackathon/infrastructure/blob/main/docs/facilitator-guide.md
- All 🤖 issues
- Report issues or bugs here: New 🤖 issue