 Hello everyone. So my topic today is unlisting the potential IoT wireless device with Kubernetes at edge and in this talk we will explore the standard approach to product project for dynamic management and observability of IoT and Oath device in heterogeneous wireless setup and we do into the region behind this approach for achieving interoperability and exploring signal integration between the different IoT wireless technologies and I am from SADU ISDRC, Internet Infrastructure Research Center from South Korea. We are working on open source by cloud native cloud management and networking and service management in an edge cloud environment. To begin with, FrothTie is a project developed by open source RedHus and it to many container work mode on small footprint device found on the edge of network. S-Divide interacts with the S-Tie cluster via S-API and then work mode is scheduled by all the labels to deploy to the device and first FrothTie managed edge device from a single point using Kubernetes CRG to handle on a kind of device configuration and edge device can register, put configuration and keep sagging the data through the FrothTie API in the edge cluster and then it's also possible to schedule container work mode to run in the device using the work mode CRD, device agent take care of running container work mode but at the same time it also take care of edge device observability by collect and put the magic to the edge side cluster. However, we have a problem with tiny things on edge for example, we have, as you know, no MCU sensor, we call it tiny things because it too small to run Kubernetes on edge due to the risk constraint and therefore we have to figure out how we can connect and manage the device to our cluster with Kubernetes and how to manage tiny things on a different variable of communication protocol. It's portable to the requirement and this is our motivation to deploy in this way and restoring and managing the device manually is not feasible at least require human intervention, especially in the large scale of deployment, it requires well-defined automated reasoning process and the second line is manually deploy workloads or updates to the end nodes is also not feasible and landing task and diverse wireless technology device and IoT protocols platform bring issue in interoperability so we deploy by to do that we use a Kubernetes CRD and a controller leveraging for our project to automate world-wide deployment also automate registration and management IoT tiny things in the interoperability and heterogeneous from for edge device is an stable Istanbul framework to adapt more IT wireless technology. The end node how to config CRD is standard from a prototype custom resource and the tiny things connect to the edge side by wireless technology for example, Wi-Fi and BRE through device block in modern here that support synchronized multiple type of connection protocols into standard structure for divide registration and configuration in the core side. Divide registering and configuring process are automated by running just only a node how to config custom resource we have developed here and for that as well for connection establishment from tiny things to edge tie with specific connection tie in the divide block in configuration and also then then the edge workload automate deploy into that tiny thing and in our implementation the edge device register corresponding with the end node how to config and with several divide property you can see in the right side and there are two work not running in this divide and finally the process of divide registration and configuration and work not deployment is automated by running end node running end node how to config with the human intervention that study our extended automation platform prototype project and all right that's how you are very soft to expand on things so we enough to have the chat with you and feel free to contact me or be too thank you for the time today