Evaluate Weigh the pros and cons of technologies, products and projects you are considering.

IoT data management gains an edge

Edge computing, the concept that data processing and IT equipment resides physically close to where data is being created, is gaining tremendous momentum in the IT landscape today. This technology is becoming widely adopted, and its use will only increase as more become familiar with its advantages. The types of organizations that benefit are the ones that have numerous locations, or edge sites, including retail stores, hotels, oil rigs, wind farms, hospitals and government agencies. Today, the explosion of IoT is changing the game as massive amounts of data from multiple devices drive edge adoption.

The days of the “big data center,” with massive processing capability and network connections to a company’s smaller sites to send data for central processing, are coming to an end as more data is being created outside the data center than ever before. In many cases, it is either too cost prohibitive to send all this data to a data center or the applications cannot handle the latency and need local, real-time processing. In fact, Gartner estimated that the amount of data created and processed outside a traditional, centralized data center or cloud will grow from 10% to 75% by 2022.

The vast quantity of IoT data being gathered at these edge sites is just too massive to be sent back to a centralized data center for processing and analysis. The world of “real-time” analysis is upon us and growing exponentially. Many IoT applications are impacting the larger IT world and contributing to today’s data explosion. Examples include:

  • Factory control systems in airports, which are responsible for temperature control, passenger information and efficient baggage circulation;
  • Facial recognition software, used by retail chains to either track criminal activity or pinpoint buying habits;
  • Traffic light systems designed to monitor traffic flow and self-adjust in real time to improve traffic flow;
  • Hospital patient monitoring systems, where doctors need access to data immediately, and there is no time to send data somewhere else for analysis; and
  • Looking ahead, autonomous vehicles are expected to generate a staggering 1-2 TB of data per car per day. That data must be analyzed and processed locally and immediately. For example, to identify traffic patterns or road blockages just ahead, the vehicle must “know now” to stop for passenger safety.

In the above examples, IoT data is being generated and collected locally, then processed and analyzed instantly. Consider the physics of the speed of light: No matter how fast a network is, the fastest data could ever travel would be the speed of light — and that simply isn’t fast enough with our current massive data sets. Even with the fastest networking gear available, due to time constraints and the need to access these data sets in real time, IoT data management must happen at the edge.

Edge computing is perfect for IoT data management and processing for a number of reasons. Edge systems are simple to administer and don’t require a dedicated IT administrator at each location for deployment or management. Since floor space is typically at a premium at edge sites, small yet powerful IT systems are required and available. Even though an organization may have multiple sites to provision, they can often be deployed and managed remotely through centralized management. Consider $2,000 saved per site — the cost savings annually becomes astronomical when you factor in many organizations that have hundreds of sites.

Edge systems offer the performance, uptime, small footprint, ease of management and low cost required for IoT data being created around the clock. Edge computing makes the complex simple for IT generalists managing IoT data.

All IoT Agenda network contributors are responsible for the content and accuracy of their posts. Opinions are of the writers and do not necessarily convey the thoughts of IoT Agenda.

Data Center
Data Management