What Does Enterprise IT Infrastructure Look Like In The IoT Age?

By Marcio Saito, CTO, Opengear

It’s very much still the early stages, but make no mistake about it: the Internet of Things (IoT) will be a permanent, ubiquitous fixture within our everyday lives, providing us a new world of connected and intelligent devices. It’s changing how we interface with (and think of) technology, how we access information, and how we put that information to use. But harnessing that information doesn’t come automatically.

As the number of IoT-enabled devices spreading across our homes, offices, and beyond continues to multiply, those connected objects gather unending data that must be transmitted and processed (often in real-time in order to deliver value). The IT infrastructure that bears this data load must be carefully considered and planned well ahead of time. The question, then, is how much of an IoT network’s data handling and processing muscle should be centralized, and how much should be placed at the network edge closer to the user?

As we peek into the next few years and try to anticipate where IoT technology will take us, the reality is it’s easy to get blinded by the hype — especially with a subject as exciting and potentially transformative as this. It’s also common for companies to miss the real opportunity that arrives after the hype quells and the real work kicks in. Maintaining a clear-eyed approach and taking the long view is what differentiates successful companies when it comes to adopting new technology wisely, a strategy that certainly applies with the IoT.

For an example of how IoT infrastructure should be considered, imagine a retail store and all of the technology and connectivity required to operate it. Not too long ago, retail was a much less technological business. A store’s entire IT profile may have consisted of the point of sale cash registers in each checkout lane, a few desktop computers that management could use to oversee inventory, and some security cameras. Each of those devices would connect to a local data closet with servers running applications to handle those customer transactions, inventory management, and video storage needs. As far as external connectivity goes, that local data would be batched up and uploaded to a corporate network server, perhaps only once a day or even once a week.

Contrast this with the retail store of today. At many locations, “cash registers” are now tablets with an attached credit card reader, running a point of sale app. Managers can access inventory data using scanners or handheld devices. Web cameras now provide security. PCs and local servers have gone away, as the devices that operate a retail store can now be controlled through apps and can connect via Wi-Fi to servers in the cloud.

Extrapolating this trend to forecast the impact of the IoT on IT infrastructure needs, we see a clear centralization of networking resources as the data load is handled less at the network edge and more at the core. Devices at retail locations are now fully dependent on internet connectivity to operate, making network resilience a much greater priority. This suggests that, as enterprise IoT technology arrives, retailers will need to focus on more robust connectivity strategies and on implementing failover capabilities to keep systems online when failures do occur (and they will — some things won’t change).Without connectivity, that critical IoT and retail technology won’t function, and the store won’t be able to conduct business.

The more important variable here, though, is the sheer size of information the IoT creates. The trend of increasing volumes of data gathered by enterprises across virtually all industries will only increase with the arrival of the IoT — and likely exponentially so.

For an enterprise-related example of how this data will be both created and consumed at the network edge, imagine every price tag within a retail store as an intelligent IoT device that can be changed remotely. Each product on the shelf will transmit data to automatically serve inventory management as well, while other sensors and beacons will monitor every aspect of the retail environment, from customer movement to purchasing behavior to security.

Considering the intensive data load this will place on the network — and that the real-time nature of this data means low-latency responsiveness will deliver required value — IT components at the network edge may be essential to getting the most out of enterprise IoT devices. With network resources such as mesh wireless gateways, data aggregators, local data storage, and analysis engines deployed at the network edge, data serving the most time-sensitive tasks can be filtered, sorted, analyzed, and put to use without needing to be sent all the way to centralized systems.

Large-scale IoT setups may even require nearby infrastructure functioning like small data centers in order to properly meet data handling demands. This IT infrastructure at the edge will likely need to rely on technologies like remote provisioning and out-of-band management in order to control costs and maintain effectiveness. Intelligent automation will play a huge role in managing the vast scale and complexity of this more disperse infrastructure built to serve the IoT.

As I’ve said, IoT technology is still developing, and taking the long view will provide the better perspective. Rather than jumping on the latest hype in the pull towards either centralized or edge infrastructure, enterprises should focus on laying the fundamental groundwork that will serve a business well under any scenario. Enterprises should look at adopting cloud and mobile technologies, and at tactics to bolster their network resilience that will serve them well now and in the future. When the IoT does drive greater needs for infrastructure capacity, enterprises that have made these preparations will be ready to grow their networks in the right direction.

Marcio Saito is CTO at Opengear, a company that builds remote infrastructure management solutions for enterprises.