Cloud and Edge Computing: Take Your IoT Solution to the Next Level

If, like most companies these days, yours has waded into the deeper water of using IoT solutions to improve business operations, then you should consider optimizing your cloud solutions to include using edge computing. Being able to leverage every advantage available to you can be a differentiator for your business operations and could dictate the ultimate success of any IoT deployment. Having closer proximity to your IoT device data can create new levels of performance for your deployments and save time (which equates to dollars) in having that data ready for real-time analytics. But, before we look at how utilizing edge computing might improve your IoT performance, let’s consider the fundamental differences that exist between cloud and edge solutions.

Initially and in simple terms, cloud computing was created to provide a cheaper and more flexible way of managing big data. By removing the operational expenses of having an on-site data center, it allowed for a budget-friendly way to compute large amounts of data or store data remotely. While it was conceived as a generalized platform that also could interface with purpose-built systems, the reality is that the data and computing is done in undisclosed locations. The shared resources of the cloud could be located in the next city or across the world. In truth, all the cloud’s resources can be routinely moved for load balancing purposes across various remote facilities. Originally, the cloud wasn’t designed to support real-time applications and due to the geographic diversity of the locations of your data, you will have longer latency when trying to retrieve that same information for analytics.

When looking at an IoT deployment, it’s clear that cloud computing is not an ideal solution for the requisite real-time feedback and immediate nature of most business IoT applications. This especially holds true when using IoT sensors in a manufacturing or supply chain environment – where the data response rates need to be extremely fast and allow for the data to be immediately available for use. The concept behind Edge computing addresses the concern of having data moved to remote locations or long distances from the IoT device to minimize latency and improve reliability. Having unreliable network support can greatly minimize the effectiveness of real-time IoT needs. Edge computing makes the transmission of the data through the various levels of media almost negligible and with a reduction of switching can improve the reliability of the network. The general location of edge resources is fixed and unmoving which results in a reduced path length between the IoT device and the edge, which by default can reduce jitter. There are a couple of edge options to understand: cloud edge happens when extending the public cloud to multiple point-of-presence designations and device edge utilizes customized software that can imitate cloud services already running on your existing hardware.

Something that should be considered when determining your system architecture is knowing exactly how you will be using the data your IoT devices are capturing. If your IoT deployments are varied – with some of them capturing data that will be stored for deeper analysis and some data that will need to be immediately ready to use – then you will need a solution that addresses all those needs. Other considerations that should be included in the implementation are fully understanding who needs to use that data and how complex the nature of the captured data might be. Having a clear picture of these details and being able to prioritize them based on your corporate strategic initiatives will provide a guide your selection. Most often, it will be a combination of cloud and edge solutions that can give your IoT deployments the full range of support requirements needed.

While both cloud and edge computing might be necessary components of your overall IoT solutions initiative, it’s critical to go through a process of detailed evaluation before making any kind of investment in implementation. Ensuring that you define how both cloud and edge can be used, possibly in tandem, to fully support your business needs is key to making sure your IoT deployments are providing the expected ROI. I recommend first determining which of your device applications need to live at the edge and perform real-time processing, immediate analytics, decision support, etc., and which applications can use more traditional cloud solutions for long term storage, post-processing analytics, etc. Based on that in-depth analysis, you can fit the most appropriate network support requirements into the overall business strategy for a complete IoT deployment structure.


  • As Director of Training and Development, Christine is responsible for the development of AOTMP University training, eLearning and the Telecom Management Tools database. She has 10+ years of mobility and telecom experience. She serves as an AOTMP subject matter expert in the categories of emerging technology, CDM, UC and cloud. Prior to joining AOTMP, she held a sales management position at Ingram Micro Mobility.

2018-07-23T14:16:34+00:00Categories: Perspective|