Evolving Edge

Organizations collect a massive amount of data from mobile and other connected devices. With advanced analytics, this data can be used to solve a number of things, such as quality control, employee safety, shopping behavior or customer experiences.

Done correctly, edge computing can even help you address key organizational and IT objectives by delivering and processing data where it’s needed – closer to applications and end users. With the edge, you can empower your decision-making capabilities through advanced analytics, deliver exceptional experiences to workers and customers, maintain cost controls and future-proof your business.

As more organizations aim to take advantage of the edge, however, it’s important to stop and examine your use cases. If you approach the edge with good planning, not only are your uses case likely to be more successful, they’ll also be able to scale effectively and efficiently without breaking the bank or your infrastructure.

You’re driving to an important meeting, when suddenly, you have to pull over because of a flat tire. You get out of your car and look at the tire. What do you do next? You might think the answer is to replace the flat tire, but why not call an Uber and get to your important meeting instead? You can call AAA from the Uber.

I call this the “flat tire syndrome.” When you’re focused on the problem directly in front of you, it can be difficult to see past that.

Most of our customers are focused on a single use case when they’re looking at different tools and devices to place at the edge. The problem you may find yourself running into with this approach is that your infrastructure won’t be able to grow with you down the road.

When you’re ready to expand your edge computing capabilities, there are four things you need to account for: flexibility, scalability, security and supportability.

Let’s say a company wants to install cameras to monitor shopping behavior. If we installed 40 cameras at this store, that could be gigabytes a day of data. You’ve also got to have high-performance analytics to process that data for real time alerts. Then suddenly, you find that what you’ve built isn’t scalable because it wasn’t planned that way from the beginning. What needs to be considered during the initial planning stages is, “In what other ways could the same infrastructure assets be used?”. For example, if you deploy video cameras to capture customer shopping behavior, and you’re processing that data at the edge, do you want to use the same edge gateways and possibly the same cameras to provide line queuing control, theft detection/shrinkage reduction, security, employee engagement for staffing decisions or other capabilities? Some of these may be funded by other departments even, such as marketing, but without considering them up front, it’s more than likely your edge deployments won’t scale to add them later.

Before you get started, here are some of the questions you need to be asking:

• Why am I doing this?
• Where does the data need to be processed? Stored?
• Should it stay at the edge?
• How much data am I collecting?
• How much needs to be centralized? How will we get it there?
• What type of storage do I need, and where?
• What data do I have to keep for compliance or security reasons?
• What data can I dump to be more efficient and effective?
• What else will I do with this platform?

A lot of planning needs to happen to avoid throwing away IT architecture down the road because it doesn’t fit the big picture.

The complexity can often appear overwhelming, to the point where you’re paralyzed by the number of moving pieces. This is why setting up a workshop assessment with the right partner is important. They’ll have experience in edge computing, and they can help you to better understand how to navigate the project and what questions you should be asking.

It’s a common misconception that a workshop assessment will slow your project down or cause delays and cost issues, but that couldn’t be further from the truth.

If done correctly, an assessment will help you accelerate the project, enabling you to get more buy-in from company stakeholders. You’ll also run into fewer issues, without missing important details that would have prevented you from implementing a flexible, future-proof infrastructure. The right partner will also help you to build a solution and test it first.

Story by Scott Davis, whois in the Enterprise Architecture practice with CDW, located in Denver. Scott has more than 37 years in IT architecture, business leadership and delivery across multiple disciplines, with a focus in data and analytics for enterprise organizations. His past roles include developer, network engineer, IT director, director of data & analytics, sales, sales management and CTO of a public company. He has worked in the retail, manufacturing and education fields, started several companies and spent time with large organizations such as Cisco, Hitachi and global integrators to develop innovative business solutions.