Avoguard blog

Introducing EdgeFlux™

What happens at your company when something goes wrong in your IoT data pipeline?
Do you have a single pane of glass view of your IoT networks for your IT and OT teams?
Do you have an integrated way to quickly and easily build, connect, monitor, and operate your IoT network and data pipeline?

If you answered no to any of the questions above, you need EdgeFlux.

By Vit Prajzler

IoT data pipelines are hard.
Hard to maintain.

Many things can go wrong on the treacherous path from the physical sensor to the final digital resting place of the data it produced. On that journey, you'll come across mechanical connectors, batteries, power supplies, wireless and wired networks, 3rd party networks, switches, routers, firewalls, load balancers, MQTT brokers, and many other hurdles.

In a complex system like that the question is not what if something breaks.
It's just a matter of time until something breaks.
The question is, what happens at your company when something goes wrong in your IoT pipeline?

Let's ponder that for a moment. At your organization or in your application,

  • How long does it take to learn about an IoT data pipeline outage?
  • How long does it take to recover?
  • How long does it take to get to the root cause?
  • What is the impact of losing the sensor data, per every minute that the pipeline is broken?

Are you happy with the answers? If not, EdgeFlux might be something for you.

Is EdgeFlux for me?

The companies we help at Avoguard have one thing in common: The cost of finding an IoT data pipeline problem is significantly higher than the cost it takes to fix it. Fixes tend to be simple. It's the figuring out what to fix that is hard. But it doesn't have to be.

When something breaks, it tends to result in tense, stressful situations for the responsible IT and OT teams. That further increases the risk of human errors. But it doesn't have to be that way.

So I asked myself,
how can I help IT and OT teams find and fix IoT problems faster and easier?
How can I help teams be more confident in these tense, stressful situations, limit miscommunication and errors?
Is there anything I can do to help prevent the data pipelines from breaking in the first place?

How to make fixing IoT problems faster and easier

This is the list I came up with as a solution to all the problems above:

  • Single pane of glass view
    • across the end-to-end IoT data pipeline
    • for end-to-end configuration, including 3rd party hardware and software
    • for end-to-end monitoring
  • It's not 3 different panes of glass, it's one screen, one interface
  • Drill down and drill up, so that it works for any scale
  • Let users tap into / inspect individual data “chunks” at any point of the pipeline
  • Build for retrofit: Add to existing infrastructure, do not replace it

With a product like that, teams would always have a real-time and historical overview of the entire system, without being an integral part of the system.

It would provide both the view of the problem, and the means of fixing the problem.
The end-to-end monitoring would make it easy to find the root cause.
The end-to-end configuration would provide means to fix it. Faster, easier.

There would be less chance for miscommunication - everything is on one screen, one display, one interface. Drill levels give each team member the right level of detail. It could do a lot of great things for both the IT and OT teams.

My mission with Avoguard's new product - EdgeFlux - is to make this reality.

Introducing EdgeFlux™

How do I plan to accomplish all of those ambitious goals? Innovation and integration. Well, at least that is the start. Apart from that, these are some foundational principles that EdgeFlux is built upon:

  • Separate data flow paths from data flow controls paths
    (software defined flows)
  • Make all interfaces observable across hardware and software
  • Integrate existing systems instead of replacing them

That's no easy goal, but solving hard IoT problems is my passion. I'll write more about how we do what we do in future posts. For now, let me say that EdgeFlux works by combining the following into one product:

  • Large scale distributed system
  • Software defined networking (SDN)
  • WebAssembly (WASM)
  • Visual Flows for great user experience on large screens
  • Mobile phone UI for great user experience on small screens
  • Infrastructure as Code (IaC) for modern IT teams and DevOps teams

* If you feel like this is just a bunch of buzzwords, please follow us on LinkedIn and Medium so you don't miss on the deep dive into the individual topics.

https://edgeflux.avoguard.com/iot-example/input EdgeFlux™ | Avoguard DEVICE Wi-Fi wise4250.advantech.local IP GATEWAY Wi-Fi gw.advantech.local IP GATEWAY LoRaWAN lorawan.gw.avoguard.com IP DEVICE LoRaWAN Motor #1 Health Sensor DevEUI CA:FE:BA::D1:BE:EF MQTT Broker mqtt.hivemq.local IP | v3.1.1 DEVICE NB-IoT my.device.tmobile.com IMEI 35-305068-412417-3 MQTT Broker emqx.mqtt.edgeflux.net v5.0

EdgeFlux UI snapshot | Input flow example

https://edgeflux.avoguard.com/iot-example/output EdgeFlux™ | Avoguard MQTT Broker mqtt.hivemq.local IP | v3.1.1 MQTT Broker emqx.mqtt.edgeflux.net v5.0 MQTT Subscription site/area/line/cell/# QoS 1 BOXLET Transformer Binary to JSON v0.3.11 @12 msg/s, 65535 msgs today BOXLET Aggregator 1 per hour v0.1.4 FILE Share file://archive.local/JSON/ Pattern {date}/{hour}.json BOXLET Transformer Binary to Influx V2 v0.3.11 INFLUXDB Client v2.influxdb.local Bucket EDGEFLUX

EdgeFlux UI snapshot | Output flow example

Challenging status quo
from the shoulders of giants

We are starting from scratch, but we have tons of role models, inspiration, and lessons learned. We are standing on the shoulders of giants such as Node-RED, NiFi, and Telegraf.

But we are taking a radical new approach.

EdgeFlux is not primarily for software engineers. We do welcome them with open arms, but they have plenty of IoT platforms already.
At Avoguard, we are building products for IT teams, OT teams, and future proofing their eventual adoption of DevOps and IaC. EdgeFlux is for them.

We are building on our experience and lessons learned in large scale wired and wireless network operations, IT systems operations, and industrial systems operations.

But rest assured, we don't want you to be starting from scratch or building on top of unproven bleeding edge technologies.

We are building EdgeFlux for retrofit.

Innovation built for retrofit

EdgeFlux does not replace existing systems, software, or hardware. Instead, it augments them.

We will help you do more with existing tooling, existing hardware, and existing software. Leveraging your previous investments, not replacing them. Replacing existing infrastructure is costly, and is the exact opposite of our mission.

Our goal is to limit the cost and risk of introducing EdgeFlux to an existing infrastructure. If you want to test EdgeFlux, adding it to your system will be as easy as removing it. You can even run it in a read-only mode for zero risk deployment.

How to get started?

Please, try it out!

If you want to increase the visibility and maintainability of your IoT data pipelines, we would love to hear from you.

For evaluation, EdgeFlux is available as a multi-tenant cloud service or as a hardware appliance on both commercial or industrial grade hardware. Sign up for an account today, or contact us to order hardware.

For production deployment, we have developed EdgeFlux so that it scales from resource-constrained edge gateways up to private or public cloud environments. You choose your preferred deployment model. Contact us to get a quote for your preferred environment.

What are the next steps for Avoguard?

Sign up and stay tuned, we'll keep you posted. We are going to be periodically publishing more about how our products solve challenging IoT problems, along with more details about the product features and solutions for specific verticals.

You can also follow us on LinkedIn, or follow me, Avoguard's founder, on Medium and LinkedIn.