Do You Need Edge Computing to Unleash IoT?


With apologies to the prolific and prescient writer Michael Lewis, there’s a new new thing, and it may surprise you to learn that it’s not the Internet of Things (IoT). 

There are two reasons why. First, IoT is now mainstream. You’ve all heard the predictions: 20 million connected things, 50 million connected things, hundreds of millions of connected things. But let’s put predictions aside and look at the reality. Organizations are already spending real money on IoT projects, and IoT is already paying dividends for many projects.

Instead, the really exciting catalyst for innovation and transformation in the IoT space is edge computing. And while you may have heard of edge computing, you may not fully appreciate that you need edge computing to unleash the full potential of IoT.

Like many of today’s new new things, edge computing came about in part because of the proliferation of cloud computing. 

I first became aware of edge computing about seven years ago, when I worked on a project in the agricultural market to send IoT sensor data to the cloud. One thing we’ve all learned about the cloud is that it’s very good for storing lots of data, and IoT certainly is driving the creation, storage and management of tons of data. But I learned a lesson about storing IoT data in the cloud that I wasn’t expecting: There was so much data that the data storage bill with our public cloud service provider was $78,000 in the first month, and we didn’t even have the project fully implemented at that point. 

That’s when I knew we needed another approach to processing and storing IoT data. And the big factor influencing our strategy was the realization that we didn’t need to act on all that data in real time. We just needed trend lines, such as humidity levels. If we knew that our levels were out of prescribed boundaries, we could act on it as necessary. And we didn’t need instanteous access to all that data to do that.

Edge computing allowed us to not just capture data, but also to analyze, synthesize and summarize it. Sticking a cost-efficient, lightweight (yet highly secure) edge system between the sensors and the cloud enabled us to reduce the need to backhaul all the data to the cloud, thus cutting costs and improving efficiency.

Another key benefit? Improved cybersecurity. Adopting an edge computing approach creates an overlay network with deeper and broader security coverage that better aligns with a more complex, diverse and demanding compute and storage infrastructure. It allows IoT systems to talk to the edge device, to other devices on the VPN and then back to the cloud. 

As organizations continue to wrap their arms around edge computing and evaluate how and where to deploy it as part of their litany of IoT projects, it’s important to keep in mind how integral edge will be not only to your compute infrastructure, but to your business processes. 

Too often, people see edge computing as just another compute cycle, and that they don’t realize that if their edge systems are down, so is their manufacturing system, or their R&D workload or their field service operation. In the era of connected things, edge computing becomes a vital cog not only in your network, but in your entire work cycle. 

It’s a pivot point in your network, especially as you move more activities to the cloud (and back to your on-premises systems, as needed). And that means you absolutely must include edge computing and edge devices in your security posture. So, as we’ve discussed many times in this space, “shifting security left” in transitioning DevOps to DevSecOps is essential, and edge has to be part of that strategy now.

At the end of the day, there are a few key things that senior business executives and board members must keep in mind about edge computing, particularly as it relates to IoT:

  1. Consider how and where to use edge computing to help deal with the complexity, cost and risks of collecting more data.
  2. Business leaders need to think about how to apply edge computing to their core business requirements, such as faster inventory turns, improved customer experience and transcendent digital transformation.
  3. There are critical security issues that must be considered as you insert edge computing into the midst of your infrastructure, architecture and business processes. Address them from the start to improve your overall security posture without having to throw more and more money and people resources at problems after the fact.

End Points

  • IoT projects need to consider how and where to use edge computing to help deal with the complexity, cost and risks of collecting more data.
  • Edge computing should be part of your overarching strategy for cybersecurity from the start.
  • Business leaders should look at edge computing as a business capability, not just a tactic in compute infrastructure.