Combine AWS edge compute services with your on-premises retail facility to improve both the customer experience and operational efficiency. Stream and analyze camera footage, optimize your point of sale systems, deploy targeted marketing content, and capture and manage application data.

Architecture Diagram

Disclaimer: Not for production use

Download PDF 

Details

  1. IP cameras stream video to the AWS Panorama appliance for computer vision analytics.
  2. An AWS Outposts server provides compute and storage to support applications deployed at the edge, such as the in-store point of sale system.
  3. Integration patterns are supported to integrate to cloud-based services such as publish-subscribe (pub/sub) by using MQ Telemetry Transport (MQTT) and API-based services. Amazon EventBridge streams events from software-as-a-service (SaaS) applications and can send order information from an ecommerce site for in-store fulfillment.
  4. Applications can present targeted marketing content driven by the video analytics on in-store displays or shelf displays.
  5. With AWS Panorama, you can manage appliances by deploying application and machine learning (ML) model packages. The models trained by Amazon SageMaker are compiled using Amazon SageMaker Neo to optimize them for inference on edge devices.
  6. AWS IoT Greengrass devices capture data such as freezer temperatures from sensors and publish to AWS IoT Core by using an MQTT topic. The MQTT messages can then be processed and visualized using additional services such as AWS IoT SiteWise. Greengrass can also be used to run custom code (as components and lambda functions) and machine learning inference locally.

Well-Architected Pillars

  • The proposed architecture will use edge and cloud-based services with Amazon CloudWatch alarms and logs. This is considered out of scope for this reference architecture focusing on solving the business case.

  • The architecture uses a combination of managed services that leave a large portion of responsibilities to AWS, following best practices of security including IAM roles scoped down, encryption at rest, and services deployed at the edge within a customer location. The edge services will be deployed using best practices such as AWS IoT Greengrass devices that authenticate using certificates, but the physical security of these will be the responsibility of the customer.

  • The managed cloud-based services are reliable by default; redundancy is built into Amazon SageMaker and Amazon API Gateway, for example. Requirements for reliability at the edge will need to be evaluated in line with business need. Mechanisms such as caching events and messages locally can be used in the event of a connectivity outage to AWS, and a significant benefit of implementing compute at the edge is bracing the reliance on a reliable network connection.

  • The cloud-based services scale to handle significant event and message volumes (such as AWS IoT Core), making use of managed services such as SageMaker Neo, which removes the need for manually scaling and monitoring performance.

  • Using serverless cloud-based services ensures minimum cost is incurred, and the edge services included in the architecture will help to reduce costs by processing data locally, which removes the need for high bandwidth connectivity to stream video to AWS for analysis.

  • The proposed reference architecture is using AWS Serverless services to have a sustainable approach; the multiple options for edge compute also facilitate choosing the most appropriate service, which helps to support sustainability.

Disclaimer

The sample code; software libraries; command line tools; proofs of concept; templates; or other related technology (including any of the foregoing that are provided by our personnel) is provided to you as AWS Content under the AWS Customer Agreement, or the relevant written agreement between you and AWS (whichever applies). You should not use this AWS Content in your production accounts, or on production or other critical data. You are responsible for testing, securing, and optimizing the AWS Content, such as sample code, as appropriate for production grade use based on your specific quality control practices and standards. Deploying AWS Content may incur AWS charges for creating or using AWS chargeable resources, such as running Amazon EC2 instances or using Amazon S3 storage.