Blueprint overview/Introduction

<Purpose- it should introduce what the blue print is about, industry, business use case, applications and where it sits on the edge infrastructure>

<It should be readable by a semi technical audience , e.g. product marketing, business account executives etc>

Use Case

KubeEdge Edge Service family focuses on a device/edge/cloud collaboration framework around KubeEdge. The verticals of focus have been IoT and MEC etc.

• The key component KubeEdge is a unique design from scratch of edge nodes and edge cloud, with all source code developed in the upstream CNCF KubeEdge Project.
• Type I of KubeEdge Edge Service family focuses an edge stack of the use case ML Inference Offloading on edge servers.
• This blueprint family will leverage various infrastructures. Arm servers will be supported. This blueprint family is infrastructure neutral.

Use Case

Type 1 project use case:

Type 1 use case is Emotion recognition scenario. This use case shows cloud, edge and device collaboration.

• Cloud - focus on Emotion recognition model training.
• Edge - provides a ML offloading service for inference.
• Device - preprocess the image and use edge ML offloading service to provide an emotion recognition service.

Image Added

Future Use cases:

Use User case 1: Smart road and autonomous driving scenario: KubeEdge facilitates smart road, deploy deploys MEC sites along the highway system. The smart road system is designed for the emerging autonomous driving industry. Those self-diving driving cars can leverage the smart road facility to process perception or navigation workload. 

Use case 2: KubeEdge can be helpful in logistics, especially useful in tracking product and equipment conditions in cold chain logistics. To maintain low temperature as well as control other conditions, cold chain cargo need needs to process inputs from sensors in it, and take actions in the real-time. KubeEdge enables cargo fleets on-board system systems to control and communicate the situation of cold-chain conditions.  

Where on the Edge

Business Drivers

Overall Architecture

<This could inform the non-technical audience, but now is more geared towards a more engaged, technical audience>

< Blue print's relation to Akraino generic architecture, how it relates to it >

< This section will use the Akraino architecture document as reference>

Platform Architecture

<Hardware components should be specified with model numbers, part numbers etc>

Software Platform Architecture

<Software components with version/release numbers >

<EDGE Interface>

Use case 3: Smart building scenario, KubeEdge can be perfectly used in helping the smart building local data processing. Lately, IoT sensors have been used in commercial buildings, however due to the data security requirement, most of IoT data need to be processed on-perm. Hundreds of data processing systems support different commercial tenants.  

Where on the Edge

Industry Sector: Cloud, Enterprise, Telco

Edge computing leverages edge locations to distribute application loads among device/edge/cloud. A service layer is required to bridge infrastructure platform and applications. e.g. load distribution coordination, hardware platform agnostic, etc. KubeEdge extends native containerized application orchestration capabilities to hosts at Edge. Along with other vertical domain support such as device twin at edge, KubeEdge edge service stack is geared to offer feature rich support to applications while remain platform neutral. 

Overall Architecture

The overall system contains cloud, edge and device three parts.

Cloud - Kubernetes platform with KubeEdge cloudcore deployed. TensorFlow training.

Edge - Linux OS, KubeEdge EdgeCore, TensorFlow inference, ML offloading service.

Device - Android smart camera or smart phone, emotion recognition App.

Image Added

Platform Architecture

Cloud - AWS virtual machine, 8 cores, 32 GB memory, 80GB EBS disk

Edge -

          On Prem Edge: Intel Xeon E5-2620, 16 cores, 128 GB memory, 3TB disks

          Cloud Edge: AWS virtual machine, 4 cores, 16 GB memory, 80GB EBS disk

Device - Android emulator, Pixel 2 API 29, Android 10.0

Software Platform Architecture

...

APIs

APIs with reference to Architecture and Modules

...

Hardware and Software Management

Hardware Management

Currently for this blueprint AWS Virtual Machines are being used for development, testing and CI/CD hence there is no specific hardware management to be done. 

Software Management

https://github.com/kubeedge/kubeedge

https://github.com/futurewei-cloud/kubeedge-android-ai

Licensing

• GNU/common license