Enterprise data centers. Colocators. Web Giants. The Cloud. When it comes to datacenter infrastructure, that’s where all the talk is. But is that really where all the data is? All this focus on what’s big and glamorous overlooks some of the most hardworking computing infrastructure in the industry.
Industrial automation is a fast-growing area of data transmission, interaction and processing. Individual machines have been automated for decades, but the current trend is towards automating entire processes, integrating the various components to achieve new levels of optimization and productivity. Outstanding examples of this are fully-automated container ports in Hamburg, Germany and Rotterdam, Netherlands.
These massive facilities transfer Ship-to-Shore containers between ships and rail or road transportation, thousands each day, day and night – and they do it with surprisingly little human intervention. The vehicles and equipment used to move, stack and retrieve containers are all automated. This is a world where driverless vehicles are already a reality. Operating independently from each other but with a high degree of coordination requires everything to be connected, and some serious computing power to keep track of it all. It’s networking at its best.
But it’s not the same kind of networking that most IT professionals think about. The data doesn’t need to be available anywhere and everywhere. Remote access is not an issue, at least not for the legions of automated “workers” in the facility. None of this data belongs in the cloud; it stays in its own ecosystem within the confines of the container port.
Does this mean that the datacenter controlling the port is less important that a colo facility with worldwide customers? Not a bit. The economic impact of a network outage in a port facility is enormous. Insuring continuous up time is as critical here as it is in the more glamorous enterprise data centers around the world.
And this is only one example of a burgeoning trend. Automated warehouses, city-wide traffic control systems, process plants, and a myriad of other applications all rely on ever more complex and expansive data integration – and are ever more sensitive to down time. They need datacenters, but they don’t need colocation, virtualization, or many of the other buzzwords that dominate the discussion of datacenter trends.
What do we call these datacenters that are appearing in massive numbers but are specific to their own industrial applications? Micro datacenters? Local datacenters? Data aggregation points? Whatever we call them, they’re a major portion of the worlds computing infrastructure.