IBM Cloud Satellite and Lumen Technologies Adapt Rapidly at the Edge

Image credit: iStockphoto/metamorworks

Workplace safety can come down to milliseconds. Whether enforcing rules for hard hats or masks, a system responsible for protecting a site needs to issue a noticeable alert almost simultaneously when it detects non-compliance.

Lumen Technologies and IBM developed a solution that meets the very low latency requirements of such use cases by processing all data — ingesting and analyzing it — where and as soon as it is generated. The solution uses video cameras to send images in real-time to a video management server, on which IBM Video Analytics software quickly processes each image, triggering an alert (if needed). Were the system to operate more slowly, a person at risk could already be many steps into the restricted area before being stopped.  

Managing video analytics at the edge

Lumen Technologies and IBM built a safety system with a set of three video cameras and two servers. The cameras are linked to one of the servers — the video management server — which is running the analytics software. This software receives and processes video images, identifies violations of movement rules, and triggers alerts. In production, the number of video management servers increases in proportion to the number of cameras, depending on the ratio that preserves low-latency performance.

Scaling up while rapidly iterating

On a separate server on-site, IBM Edge Application Manager runs in containers on Red Hat OpenShift for IBM Cloud; its role is to install the most recent version of the analytics software on all video management servers there. As the number of video management servers in a deployment increases, so would containerized instances of the Edge Application Manager and the number of OpenShift worker nodes needed to support them.

Consistent deployments across locations

But what is a deployment of Red Hat OpenShift for IBM Cloud doing in an edge site?

As you have already guessed, the short answer is IBM Cloud Satellite.

In setting up the solution with the video cameras and video management servers in place, a customer’s operations team first uses IBM Cloud to select hosts at the edge site to server as the Satellite “location.” Once the location is set up, the team uses the same IBM Cloud console to provision Red Hat OpenShift for IBM Cloud in that new location and deploy the Edge Application Manager in containers to pods on virtual machines serving as worker nodes.

And this is the key to scalability for this safety solution. Besides putting video management servers in place and linking video cameras to them, rolling out their solution in new sites is easily accomplished by setting up Satellite locations, provisioning Red Hat OpenShift for IBM Cloud, and deploying the appropriate number of Edge Application Manager instances to worker nodes.

The consistency of software across all locations is ensured through the single view in IBM Cloud, from which cloud services, containerized applications, security, and network policies are monitored and can be managed across public and private environments.

Adapting to emerging needs

Since the video analytics software can be trained to identify any visual pattern and enforce different movement rules related to what is observed, the safety solution is adaptable. For example, with a thermal camera for COVID-19 monitoring, retrained video analytics can allow employers to instantaneously detect employees’ temperatures. For that same use case, other camera analytics can calculate how many people are using a space and determine when the next deep cleaning is needed.

Continuous security and observability

A single, consolidated view of Satellite locations shows deployments and services running in every location. Teams can manage the network traffic and configure the applications within all locations and provision and use services as if they are working in the public cloud. That also means client teams can even deploy the same application stack to any location from the IBM Cloud catalog.

Satellite Link establishes secure tunnels and enables control of application and service traffic to and from each location. Satellite Link works with your existing network configuration and security postures. Teams in all Satellite locations use the same access and identity management (IAM). With support for a customer’s own keys and certificates, consistent data encryption enables workloads to span locations securely. Endpoints across the secure tunnels are uniquely and automatically named, yielding fast DNS, predictable operations, and easy compliance audits.

Consistent and portable operations at any scale

Lumen Technologies and IBM built a solution that can perform real-time, intelligent data analysis at thousands of edge sites across high-speed fiber connections to the many Lumen Edge platform locations where IBM Cloud Satellite and the Edge Application Manager run. Through a single view in IBM Cloud Satellite, operating the solution is consistent across all hubs and locations. That repeatability is a baseline from which teams can gain velocity in rolling-out deployments, quickly scaling up edge locations with new functionality, and remotely automating many operational chores.

The original article by Briana Frank, director of product management at IBM, is here.

The views and opinions expressed in this article are those of the author and do not necessarily reflect those of CDOTrends. Image credit: iStockphoto/metamorworks