Computing on the edge

Jos

TS Evangelist

It’s easy to fall into the trap these days that anything really important in tech only happens in the cloud. After all, that’s where all the excitement, investment and discussion seems to be. To be sure, there are indeed innumerable efforts to not only build software for the cloud, but also to use the cloud to completely reinvent companies or even industries.

As important as these cloud-based developments may be, however, they shouldn’t supercede many of the equally exciting capabilities being brought to life on the edge of today’s networks. While these endpoints, or edge devices, used to be limited to smartphones, PCs and tablets, there’s now an explosion of new options for creating, manipulating, viewing, analyzing and storing data. From VR headsets to smart digital assistants to intelligent tractors, the range of edge devices is enormous and shows no signs of slowing down anytime soon.

In addition, we’re starting to see the appearance of entirely new types of distributed computing architectures that can break up large workloads across different elements. Admittedly, some of this can get pretty messy fast, but suffice to say that many types of modern applications, such as voice-based computing, big (and little!) data analytics, factory automation, and real-time document collaboration tools all require the efforts and coordination of several different layers of computing, including pieces that live out on the edge.

On the industrial side of this work, there’s a relatively new industry group called the OpenFog Consortium—originally organized by companies like Cisco, ARM, Dell and Microsoft—that’s working to try and standardize some of these elements and how they can be used in these types of modern applications. The group gets its somewhat confusing name from the concept of applying cloud-like computing principles close to the ground (I.e., near the edge or endpoint)—similar to how clouds near the ground are perceived as fog in the real world.

Major tech companies all see connected cars as essentially ‘the’ computing device of the next decade or so, just as smartphones have been for the last decade.

In many fog computing applications, sensor data from an endpoint device or attached straight into a simple server-like computer (sometimes called a “gateway”) is acted upon by that gateway to trigger certain actions or perform certain types of tasks. After that, the data is also forwarded on up the chain to more powerful servers that typically do live in the cloud for advanced data analysis.

Probably the best example of an advanced edge computing element is a connected autonomous (or even semi-autonomous) car. Thanks to a combination of enormous amounts of sensor data, critical local processing power, and an equally essential need to connect back to more advanced data analysis tools in the cloud, autonomous cars are seen as the poster child of advanced edge computing. Throw in the wide range of different types of computing elements required for assisted or autonomous driving, and it’s easy to see why so many companies are making major acquisitions in this space.

Intel’s announced plans to purchase MobileEye yesterday, for example, is just the latest in a string of key developments in this market and it’s not likely to be the last. MobileEye’s components will bring computer vision and other critical elements of connected car-based computing to Intel’s rapidly growing grab bag of complimentary technologies.

Companies like Intel, nVidia, Qualcomm and ARM on the semiconductor side, as well as system integrators like Harman (recently purchased by Samsung) all see connected cars as essentially “the” computing device of the next decade or so, just as smartphones have been for the last decade. That’s another reason why there’s so much excitement—and so many battles looming—in and around car tech. Add in the carriers, network providers, car OEMs, other Tier one suppliers, and a raft of startups and the stage is set for an intricate and complex competitive dance for years to come.

While it’s tempting to long for the simpler days of computing devices, where everything occurred or locally, or even a pure cloud-based world, where everything happens in remote data centers, the simple truth is today’s advanced applications require much more sophisticated hybrid designs. Building out a cloud-based infrastructure and cloud-based software tools was a critical step along this computing evolution chain, but it’s clear that the most interesting and exciting developments moving forward are going to be pushing advanced computing out onto the edge.

Bob O’Donnell is the founder and chief analyst of TECHnalysis Research, LLC a technology consulting and market research firm. You can follow him on Twitter . This article was originally published on Tech.pinions.

Permalink to story.

 
Last edited by a moderator:

amstech

IT Overlord
On a consumer level and non-vital business models sure, Cloud services are growing.
In many other areas though, they are not.

Many large scale computing infrastructures, especially Medical are actually going away/staying from cloud services...atleast for certain data. Anything that is really vital will not be trusted with cloud services anytime soon. Sure, if your saving recipes or car designs that's one thing. Personal data and Medical information? Not so much.
The data is all in house [and backed up] and todays patient portals allow the customer to view their information the same way without worrying where their data is. I worked at a large Hospital in Upstate NY and now I work for the State of NY (SLPC) and we use a few cloud services for non vital services, but that's about it.
 

w8676

TS Rookie
. I worked at a large Hospital in Upstate NY and now I work for the State of NY (SLPC) and we use a few cloud services for non vital services, but that's about it.
SpiderOak and a handful of others have been HIPAA compliant for many years at this point. I think your point more shows that healthcare is ossified and bereft of competitive pressure to improve processes, not that the cloud isn't more than capable of offering that industry the benefits it offers others. If for absolutely nothing else, HIPAA-compliant off-site backup of self-hosted (Nextcloud, Seafile, Tonido, etc) solutions would strike most businesses as a good idea.
 

DWMK3

TS Rookie
On a consumer level and non-vital business models sure, Cloud services are growing.
In many other areas though, they are not.

Many large scale computing infrastructures, especially Medical are actually going away/staying from cloud services...atleast for certain data. Anything that is really vital will not be trusted with cloud services anytime soon. Sure, if your saving recipes or car designs that's one thing. Personal data and Medical information? Not so much.
The data is all in house [and backed up] and todays patient portals allow the customer to view their information the same way without worrying where their data is. I worked at a large Hospital in Upstate NY and now I work for the State of NY (SLPC) and we use a few cloud services for non vital services, but that's about it.

Do you think that with time though the vital services will be moved to the cloud as cloud computing costs are decreasing as well as the security with cloud services improving?

I know that uptime would have to be near 100% for something like a medical center so maybe these types of businesses will always use in house solutions?
 

Uncle Al

TS Evangelist
I'm just not sold on this centralization concept of the cloud, especially in these times of hysteria, hacking, stuxnet knock-offs, etc. I use it strictly as a 3rd back up but certainly won't use it as my primary until they hold mass executions of hackers in Times Square( or maybe at Trump Tower?) each Monday so we have a real deterrent with measureable results ...... Oh, don't forget the guy selling beer and peanuts!
 

Yynxs

TS Maniac
. I worked at a large Hospital in Upstate NY and now I work for the State of NY (SLPC) and we use a few cloud services for non vital services, but that's about it.
SpiderOak and a handful of others have been HIPAA compliant for many years at this point. I think your point more shows that healthcare is ossified and bereft of competitive pressure to improve processes, not that the cloud isn't more than capable of offering that industry the benefits it offers others. If for absolutely nothing else, HIPAA-compliant off-site backup of self-hosted (Nextcloud, Seafile, Tonido, etc) solutions would strike most businesses as a good idea.
I am always amazed at statements like this. From a security standpoint alone, there is nothing in or on the cloud guaranteeing safety of anyone's medical information. Nothing. The very EULAs read "best practices" to blunt any legal generalities about systemic failure in class actions.

Retaining patient data on local servers a) prevents targeting by the importance of the individual, which every individual, important or not, believes about themselves let alone government or public interest individuals; b)requires more 'effort' on the malefactors to get large volumes of data and repeating their efforts increasing the chance of random discovery; c) retains some measure of privacy for the individual with local permission information sharing controls available; d)individualizes the fail point should data loss occur.

I am offered "access" to my medical information on the web by every provider these days. I turn it down and tell them delete any web passwords, standard or otherwise. If breach occurs, the local provider then cannot say in court, I 'MAY' have been the weak point. If my data goes to the 'cloud' it wouldn't matter, I as the consumer and victim now must prove where the weak point occurred often outside state lines. In short, the 'cloud' destroys my ability to seek redress for stupidity because "best practices" could be defined in California where federal judges can be bought and sold (see Google et al. wage suppression decision).

To your point, health care is not 'ossified'. It is pointed towards and highly aware of the welfare of the user (called 'the patient'). No knowledgeable patient wants their information in the cloud. "oooh! LOOK! SHINY!" people believe the patient is somehow gaining something. The patient is not. Companies and malefactors wanting to exploit the cloud-based storage do. There is no cost savings to the patient between a local data backup and a cloud data backup. There is an incredible security threat increase between local and cloud.
 
  • Like
Reactions: amstech