A Blog by Jonathan Low

 

Jul 2, 2020

How Covid Is Accelerating the Internet of Things And the Need For Data Storage

More people working from home on more devices demanding more digital control means exponentially greater storage needs. JL

Venture Beat reports:

IoT devices have become a big part of the new reality. They’re making it possible to limit dense gatherings of workers to avoid virus transmission, for example. And interwoven in the rise of IoT is an even stronger demand for processing and storage.It’s particularly a problem when crucial business video conferencing requires HD video. The pipes aren’t large enough to do that with acceptable performance. (But) very little of that data is worth storing (though) increased human and machine communications during a pandemic are centered around data that has to be created, stored, and shared.
In the midst of the pandemic, IoT is no longer a nice-to-have but a need-to-have. Many people and businesses are relying on IoT products such as remote connected health monitoring solutions, packaging and shipping trackers, and streaming devices — the devices that are enabling remote work, telehealth, and telelearning. It also means the amount of data being transmitted, received, stored, and analyzed at the edge, or on devices, is exploding.
“For many people, before COVID-19, IoT was a nebulous thing, devices talking to devices,” says Huibert Verhoeven, senior vice president and general manager of Automotive, Mobile and Emerging Markets, Western Digital. “Now these IoT devices have become a big part of the new reality. They’re making it possible to limit dense gatherings of workers to avoid virus transmission, for example. And interwoven in the rise of IoT is an even stronger demand for processing and storage.”

The need for latency-free data transmission

As we continue to see where the pandemic leads us, home internet service providers have realized how critical data is and have had to plan for increased at-home use of systems — in some cases, suspending the caps on their customers’ home internet plans.
It’s particularly a problem when crucial business video conferencing requires low latency immersive HD video. The pipes simply aren’t large enough to do that with acceptable performance. At the same time, very little of that data you send and receive is worth storing, and most of it only has value for a small period of time. Of course, some of that data is actually tremendously valuable, but might require added AI software to extract it — and of course that’s immensely compute intensive.a
Companies need a strategy for data management. In some cases, data is used and analyzed immediately, and in other cases it is used later, or simply retained (think of regulations on record retention), and in other cases, it is stored locally, as system architects realize that some data needs to remain close to where it was captured.
The aim for companies working with a vast array of IoT devices is to place specific storage solutions where they are most needed to ensure that data is handled appropriately across its entire journey, and that the user experience in examples like HD videoconferencing maintains a high quality of service.
Privacy concerns also mean that some end users or companies might prefer that their data stay local, that is, to be analyzed and processed on the device rather than in the cloud. This scenario tends to favor placing storage on-premises in common edge applications like smart video in cameras using embedded, or even removable storage card solutions.
However today, data is not just stored at the edge, and there is the challenge of managing the distributed nature of data.

Data demands placed on distributed hubs

Data today is much more distributed. Data could be located in edge data centers, or in any of a number of smaller distribution sites. It’s also locally stored on devices and can be locally analyzed on devices too.
Consider the data demands of distribution and fulfillment: they’ve been massively disrupted because of the impact the pandemic has had on shipping conditions and distribution centers. Before COVID-19, typical requests would go to a central point where they would be disseminated to the biggest hub closest to the consumer with most of the supplies in stock. The result: the two-day delivery that disrupted the industry so long ago.
But even that has gotten significantly more complex with the enormous surge in demand for shelter-in-place supplies, groceries, electronics, and so on. Massive distribution is a challenge, because of the sheer volume of requests and new precautions for high population and dense worker environments.
Many companies are now looking at how to set up smaller locations closer to the customer, which creates a supply chain issue, Verhoeven says. You would no longer have larger data centers with central data resources. Yet, all that information has tremendous value from an efficiency perspective, a market perspective, a routing perspective, and more. All that information, all of a sudden, is even more important.
“Companies are starting to realize that data is of vital importance to their survival and to growing their business,” he says. “The analytics on that data is so important that many midsize companies are now taking it in-house and deploying on-edge data centers, as opposed to outsourcing or hosting with clouds or other data centers.”
That means the need for more edge data centers for all of these customers is growing, and becoming essential. For medium- to large-size companies, deploying many more of those edge data centers means they need their own specific storage. For latency reasons, for speed reasons, it’s a fundamentally different architecture, because they’re trying to solve slightly different problems.

The move toward AR/VR and meeting virtually gets real

Another advance that COVID-19 has brought to bear is the use of virtual meetings. For companies developing new technologies or running a global business, the required expertise won’t always be in the same location of a problem that needs to be fixed.
“Before COVID, I would have support engineers fly between distant continents on two hours’ notice to make sure we can help a customer at some stage in product development,” says Verhoeven. “With AR and VR, will we see these engineers in the same lab, looking at the same thing, on a common whiteboard while working in augmented reality?”
The AR/VR space too, could get a tremendous boost from what we are going through, Verhoeven asserts, enabling this kind of immersive long-distance collaboration, taking AR/VR far beyond the gaming world where it’s mostly been confined up until now. This new virtual work environment is another example of how certain workflows can actually improve or become more efficient, if the right technology is in place.
Verhoeven says, “Today, the importance of storage is undeniable as it plays a role in all these data scenarios, at home and in business settings. It must be considered as part of one’s business strategy. Not only does storage support human and machine-to-machine communications, but when combined with AI, storage enables companies to access data quickly to gain insights. Access to such data will be important as new post-pandemic business models develop.”

Bolstering connectivity and the data infrastructure

There’s a big reason to look at the role of data infrastructure to ensure that mission-critical data can be transmitted, received, stored, and analyzed where it’s needed and when. Most important is a boost in connectivity — fiber to the curb that takes internet speeds to a gig and beyond.
That kind of internet connectivity and speed is becoming increasingly available across the U.S., and availability has begun to accelerate as demand continues to increase. Then there’s 5G.
For several years, 5G has been getting a lot of hype because users want the ability to connect anywhere, and share large data files and videos, and 5G seemed set to deliver. But while 5G phones have been introduced globally, networks have lagged, and few consumers have access to 5G.
“I think we have found in this situation a catalyst that will accelerate the demand for 5G,” says Verhoeven. “The newly dispersed workforce still requires quality virtual connections, and this will continue to drive demand for high-speed, low-latency connectivity everywhere, even on the go.”
The pandemic has shed a light on ways that 5G, were it fully deployed globally, could help home-based workers and or workers still onsite who are focused on mission-critical manufacturing and other work. More efficient and responsive storage in 5G smartphones is an area Verhoeven’s team focuses on. 5G is a key driving force in helping IoT move forward enabling more reliable autonomous manufacturing processes via new standards for ultra-low latency in factories. The processing power required for 5G is tremendous, and along with that comes the requirements for data storage.
In IoT, the processing and storage requirements needed for factory endpoints such as robots and cameras are growing. In autonomous driving, many latency- and privacy-sensitive processes running in the car will require careful architectural storage design across many subsystems.
Storage technologies provide many options today. Companies can host their data in the cloud, or control their own destiny by implementing scalable storage solutions and infrastructure for their business, on-premises, at the edge, or at the endpoints. High-capacity hard drives enable the cloud, fast NVMe™ SSDs propel data in the data center for AI, and embedded flash storage at the endpoints enables on-device data analysis.
“All these examples of increased human and machine communications during a pandemic are centered around data that has to be created, stored, and shared to bring results and insights,” explains Verhoeven. “Whichever storyline — WFH, smart devices, telemedicine, supply chain — whichever thread you start pulling, it all leads to more data, going to more places and growing in its potential value.”

0 comments:

Post a Comment