Edge analytics—the approach to data collection and analysis whereby automated analytical computations are performed at a sensor, network switch, peripheral node, or another connected device, rather than later, after sending the data to a centralized data store. The analysis takes place in near real-time at a non-central point, in a “decentralized” environment.
“One way to understand edge analytics is as an alternative to traditional big data analytics, which is performed in centralized ways, through Hadoop clusters or other means, often from a big data warehouse or another central repository. This has been a popular way to drive analytics, but now, data scientists are exploring how edge analytics can work as an effective alternative option,” explains Techopedia.
This type of analysis is proving to be both a timesaving and resource-saving alternative, as it cuts out the “middle-man”—the warehouse—by delivering data directly to analysts.
WHY edge analytics?
“As it stands, the existing IoT [Internet of Things] model isn’t working. The concept of transferring vast amounts of data that are collected from an army of sensors to a central repository is neither sustainable or affordable,” advises Peter Ruffley, chairman of Zizo. “It is the ability to analyze data at the edge — effectively on site — that will unlock meaningful new opportunities for businesses . . . and undoubtedly changes the way IoT can be leveraged – and will mean that the return on the investment in IoT can finally be realized.”
Some are questioning whether edge analytics is but another “gimmicky term” with an eventual result of complicating matters more than is necessary. While such a concern is worth reflecting upon, evidence to the contrary is easy to find; such as the forecasted growth of IoT devices to reach more than 64 billion by 2025, up from about 10 billion in 2018, and 9 billion in 2017. It’s difficult even to fathom the enormous amount of data that this vast array of technology will produce.
Ramesh Dontha, the managing partner at Digital Transformation, reminds us that “Organizations are deploying millions of sensors or other smart connected devices at the edge of their networks at a rapid pace and the operational data that they collect on this massive scale could present a huge problem to manage.”
While the insights gleaned from this data will benefit everyone from product developers to business analysts to the consumer, the sheer volume of data is impeding the analysis, resulting in a slow, inefficient process.
“Edge analytics aims to streamline the data analysis process to utilize as much of the relevant information as possible, more efficiently than traditional methods,” notes Vivian Zhang of the NYC Data Science Academy.
The WHO and WHERE
Real-time/near-real-time data insights can benefit any number of sectors. Industries leading the way in deploying data analysis on the edge are retail, manufacturing, energy, smart cities, and transportation. Specific uses for this efficient method include retail customer behavior, remote monitoring, and maintenance for energy operations, as well as fraud detection at financial locations such as ATMs and observing manufacturing equipment.
“Getting to edge analytics is not an overnight task, and it typically involves creating the analytics model, deploying the model and executing the model at the edge,” shares Dontha, “There are decisions that need to be made in each of these areas concerning collecting data, preparing data, selecting the algorithms, training the algorithms continuously, deploying/redeploying the models, etc.”
Join us next time for a look at the pros and cons of utilizing edge analytics.
At RomAnalytics, we’re connectors. As a leader in market insights and analytics talent recruitment, we work with our clients and candidates to create the best professional match. Our years of experience, a wealth of knowledge, and passionate pursuit toward every job placement result in satisfying experiences for both clients and candidates. Give our team a call today.
Leave A Comment