AWS Smart Manufacturing at the Connected Intelligent Edge

Wednesday 10/5/22 01:33pm
Posted By Ramya Kanthi Polisetti
  • Up0
  • Down0

Snapdragon and Qualcomm branded products are products of
Qualcomm Technologies, Inc. and/or its subsidiaries.

We have conversations all the time with people trying to understand the business value of combining cloud computing and edge computing. And when we show them an application like the AWS IoT manufacturing demo that Thundercomm recently presented at Embedded World, we see the light bulb go on over their head. It’s the kind of aha-moment we like to promote.

Take luggage tags, for example

Connected Intelligent Edge: Smart Manufacturing

The demo involves a simple manufactured product: luggage tags.

Of course, the product is simple, but there are a number of steps on the path to producing it:

  • Take the order
  • Convey the order to the engraver
  • Load the engraver with blank tags
  • Engrave the tags
  • Unload the engraved tags from the engraver
  • Inspect the tags for flaws or mistakes
  • Isolate rejected tags
  • Hand accepted tags off for delivery
  • Repeat

The goal of the demo is to showcase a manufacturing application that spans both cloud and edge computing.

Ordering through the cloud, manufacturing at the edge
First, the user enters luggage tag details in the cloud through a web interface. That sets up the order in AWS Cloud.

Also in the cloud, the order is stored and converted to engraving instructions that AWS Cloud transmits over 5G or Wi-Fi, as shown in the diagram below.

Click Image to Enlarge

In the manufacturing facility at the network edge, the Thundercomm EB5 Edge AI Box processes all instructions locally. It controls the engraver and the robotic arm that moves the tags to and from the engraver.

After engraving, the arm moves the tags to a Basler camera where they undergo quality inspection using AI inferencing that runs on the EB5. The arm moves accepted tags to a delivery bin and places rejected tags in a discard bin.

The EB5 Edge AI Box is built on the Qualcomm® QRB5165 robotics processor. It offers AI, machine learning, heterogeneous computing, enhanced computer vision and an image signal processor for multi-camera concurrency. The QRB5165 is suited to industrial-grade temperature ranges and offers security at every layer.

Next step: The connected intelligent edge
In the connected intelligent edge, processing stays on edge devices that remain connected to the cloud but grow less dependent on it. It’s a typical example of how the Internet of Things evolves as things become more capable.

This use case takes the decision-making to the data instead of shuttling data between manufacturing equipment and a central server. It highlights how to run machine learning inference locally for object detection and quality control and it shows how AWS Cloud integrates with partners that provide hardware and machinery.

Find out more about the Thundercomm EB5 Edge AI Box and see how you can build it into your applications. Explore the AWS Cloud and AWS IoT projects created by your peers in the Qualcomm Developer Network.

Qualcomm QRB5165, Qualcomm Adreno, Qualcomm Hexagon, and Qualcomm Spectra are products of Qualcomm Technologies, Inc. and/or its subsidiaries.