JRB - Fotolia
Several years ago, the "intelligent edge" meant designing an IT strategy to support rich computing devices such as smartphones, tablets and PCs. But endpoints today aren't all what they used to be.
Edge computing involves processing and analyzing data at the actual site where the data is created. Enterprise mobility management (EMM) software burst onto the scene to secure and manage new mobile devices. To streamline management, EMM providers added features that would converge control of PCs and mobile devices into one platform called unified endpoint management (UEM). In some cases, they added new identity management services to support linking users to their devices.
The new edge includes billions of internet of things (IoT) devices that range from sensors to airplane control systems, each with varying levels of software intelligence and processing power. Integrating these IoT edge devices into existing endpoint management strategies requires IT to change the way it designs and architects three key management systems: device management, security and data analytics.
First, these IoT devices communicate with a wide range of protocols, and they have varying types of processors and storage. In some cases, IoT edge devices have restricted hardware capabilities for processing power and storage, which means they can't run advanced security and management software.
Additionally, these newly connected IoT devices frequently operate in harsh environments and remote locations, and older devices were never designed to connect to the internet. As a result, they don't fit neatly into the existing mobile and PC management framework.
Finally, the person that manages the IoT edge devices may be an operational technology (OT) executive, such as a plant manager, instead of an IT professional.
UEM evolving to intelligent edge management
To support OT and IT device management and security needs, technology teams will need a new system that enables OT to securely onboard, provision, configure, maintain and monitor connected sensors and equipment. But it's not just OT that needs these systems; industries such as retail will also support a wide range of connected sensor-enabled devices.
These intelligent edge management (IEM) systems will collect telemetry data, analyze the data stream to determine the operational health and status of the system or device, and deliver software updates to maintain operating systems and apps. IEM must support multiple legacy and proprietary protocols for industries such as industrial automation, automotive and healthcare. In addition to connecting various devices and analyzing new data, IEM must scale to support thousands and even millions of different units very quickly.
Myriad IoT management platforms that support many of these functions have cropped up. The problem is many of these platforms weren't integrated with the existing IT stack, opening up the potential for security issues and creating management headaches. Today, the top cloud providers have entered this marketplace with new IoT platforms that are focused on solving many of these issues, and many of the original IoT platforms have consolidated. Those that still exist are looking to add more functions to support combined OT and IT functions.
Meanwhile, UEM providers have the opportunity to bridge the gap between OT and IT systems by streamlining the management and security of IoT devices. Even companies whose IoT efforts are more IT-centric need to purchase tools that vendors built with IoT edge devices in mind. Regardless of the type of vendor an organization selects, scalable offerings are available to secure and manage devices for a range of industries.
Intelligent edge needs a data management strategy
Device management and security aren't the only considerations in designing an intelligent edge management strategy. Organizations also need to define a data storage, processing and analytics strategy at the edge. In the first wave of IoT deployments, many organizations attempted to stream data from the edge to the cloud for real-time analytics.
Today, they understand that cost, latency, bandwidth and security mean that some -- if not all -- of a company's IoT access points will be managed on site at the place of creation. For example, a factory may have limited bandwidth at each of its sites, and streaming video surveillance footage to the cloud can be costly. In other cases, a company can't wait for the data to go to the cloud for processing. The timeliness of analytics, such as manufacturing line operation, may require the company to analyze information in real time locally. Additionally, a company may decide that even with encryption, it doesn't want the raw data leaving its facility for security reasons. In this case, the business will analyze the data and only send the results to a centralized data lake.
Processing at least a portion of the data at the edge will enable IT to save on bandwidth and storage costs. It can also allow for faster analytics to drive new business value such as minimizing downtime and offering equipment as a service. That's why the third leg of the stool in designing an intelligent edge strategy requires understanding where data will reside and be analyzed.
It's an exciting time in the connected device world, but organizations need new management tools and approaches to minimize security risk because IoT edge devices dramatically increase the attack surface. This is easier in organizations where IT is responsible for all devices. For more industrial verticals, OT and IT must work together to manage and secure devices and data. All companies must also learn new skills around creating an analytics strategy that supports new data and integrates this data into existing workflows. If organizations appropriately manage an endpoint strategy with IoT edge devices in mind, the intelligent edge will deliver new insight and profit.
- How to manage emerging technology in IT –ComputerWeekly.com
- Efficacy of Emerging Network Security Technologies –ComputerWeekly.com
- October ezine: Tap into wearables –MicroScope
- Emerging Technologies in Cognitive Computing and Machine Learning –IBM