Deployment for Connected Robotics

The deployment package aims to deliver an easy deployment and onboarding process to the Connected Robotics platform. It consolidates a few technologies that allow to automate the process of provisioning the required architecture in the local and cloud environments.

Infrastructure as code deployment base

Robot onboarding

Onboarding of a new robot requires an Onboarding file that may be available in guide mentioned in the link. If your robot is in this list you may use this template and just adjust it to the minor changes you may have. If your robot, however, is not on the list, you will have to generate the onboarding template before. This template is a json that needs to be included in the body of the http request. Download the script and run it in your robot: The guide for robot onboarding can be found here:

Network Application onboarding

This section shows how to deploy a Network Application to the Middleware system. Essentially, this is the execution of robotics tasks that have been previously onboarded. So, for the deployment of a network application, the taskid containing the app to be deployed is necessary. Follow the tutorial:

Cloud Edge onboarding

For the middleware to plan optimal network application placement, it is important to import the network topology into the system. This section shows how to create a new Cloud or edge entity in the Redis backend of the Middleware.

Adding a new entry in the Middleware topology allows Middleware to better optimize the resource and task planning to provide the best network capabilities to the Robot. Thanks to this, the Network Application placement can be adjusted to specific needs like low latency network thanks to the closer placement from the Robot or specific network slice requirement (upcoming!). The full guide on cloud and edge onboarding can be found here :

AWS deployment with Terraform

The Terraform package aims to deliver a simplified onboarding process of the Middleware working under the private cloud provider AWS. It automates the deployment process of the required architecture that will be ready for the Middleware deployment. The package provisions required VPC with public-private subnet architecture and deploys the Kubernetes EKS Cluster in the subnets. Additionally, the Kubernetes cluster is configured already to be ready to deploy the Middleware with all required security features.

Micro K8s deployment

The Microk8s deployment allows provisioning the Middleware in the local Edge locations. It is the local Kubernetes cluster that will allow quick onsite processing by utilising all the capabilities provided by the platform.

Wiki: Deployment for Connected Robotics (last edited 2023-06-20 11:11:07 by Diwali)