Cloud

The Internet of Things will cost companies more than they’re ready for

Above: Sen.se's "Mother" Internet of Things gadget

Image Credit: Devindra Hardawar/VentureBeat

Though the Internet of Things era has only just started, it may already be broken.

Like generals fighting the last war instead of the next one, many companies working to build the Internet of Things seem to be stuck in the smartphone and tablet era, embracing approaches that will soon be obsolete, if they aren’t already.

Today, smartphones are powerful hubs surrounded by less intelligent objects. Each device is managed and operated from a few centralized data centers. This is not yet a major issue as devices currently last only a year or two before being decommissioned. The cost of managing data centers is limited in duration and underwritten by a constant flood of replacement devices with short lives.

Not so in the Internet of Things era: an LED lightbulb has an expected life of 20+ years; aircraft are expected to remain in service for decades; the average car on the road in the US is now more than a decade old. Applying a centralized cloud-based business model to these devices will mean decades of expense without decades of associated revenue. At IBM we already see clients that are struggling with device-related services that have failed to meet revenue targets, but cannot be switched off for fear of angering an installed base.

There is a solution at hand to the mismatch of management costs and revenue expectations for the Internet of Things: distributed, edge-based cloud computing. If the data center is the center of the network, PCs, smartphones and other connected devices are on the edge. Edge-based cloud computing is about more than making individual devices smart and connected. It involves linking together every device at the edge of the network to form an integrated, distributed cloud service. If we can get these smart devices to manage themselves, then network services will be available for as long as the devices are in place and at an extraordinarily low price.

Today, IBM is developing an entirely new architecture that enables a full range of computing capabilities in a distributed model. Unlike past approaches to distributed computing, this one does not depend on knowing or trusting every node in the network — a practical impossibility with billions of devices to be placed in millions of locations.

The core of this new approach is built upon the Block Chain, a model of distributed computing leveraging the architecture of BitCoin (without the financial component). Using the Block Chain we can implement the typical transaction processing work done by centralized data centers without any of the cost associated with those systems by using compute power generated by individual devices that would, in most cases, go to waste.

These distributed, Block Chain-based services will run on new transport protocols as well. One candidate is a newly developed protocol called Telehash that offers peer-to-peer based transport and messaging that is, by default, heavily encrypted. Large file transfers can be run through the very well-established BitTorrent transport tool.

In this vision of the future where the cloud stretches all the way to the edge of the network, devices will have the ability to connect directly with each other, manage themselves and even trade services with each other, offering redundancy and back-up protection without additional cost. Putting these technologies together, we can build a far more efficient, secure, durable and lower cost Internet of Things.

Paul Brody is vice president and global electronics industry leader for IBM Global Business Services

32 comments
jg_networks
jg_networks

@S_dF well yeah because there's going to be massive security breaches everywhere

Sander ten Wolde
Sander ten Wolde

So, his business argument seems to be this: 
Applying a centralized cloud-based business model to devices that will not change for 10 years, will, over time, mean spending lots of money on legacy equipment with few users and generating very little revenue? Example I think: the in-car navigation systems that become old over time, compared to a phone/tablet based one that is software updated every week. Counter argument: if the device is 'connected', it becomes a question of the limits of the hardware versus being able to update the software. 

Rafael Kireyev
Rafael Kireyev

Attractive idea, but it is very likely that over time, the network security problem will arise. Even more likely, that this problem will entail significant expenses. And finally, the most probable that these expenses will be passed on to customers.

Ajay Desai
Ajay Desai

That's why when I go big data (in consulting and other projects) it's Zetabyte or bust.

Richard Ortega
Richard Ortega

I work on IoT and these feels like an IBM ad... Plus they're overestimating compute power compared with battery power.

Daniel Data
Daniel Data

That looks like the robot from Superjail!

Brad Chun
Brad Chun

Nick Kuyakanon Matt Sheppard

Debraj De
Debraj De

The best IoT products and solutions are the ones, where human users will have to interact the least (annoyed with seeing everything trying to turn into notification and control on smartphone, too much distractions!), people will conduct daily life without mostly even noticing presence of those special networked devices embedded in physical environment around them, and will see detectable improvement of their day to day activities and life management. That's the true IoT. :)

theRab
theRab

@andypiper that's what i thought. the internet allows finds a way...

S_dF
S_dF

@jg_networks Reframes the security conversation around encryption & response not 'castle wall' protection

Paul Brody
Paul Brody

We selected the BlockChain and the concept form BitCoin because we think that, over time, the strong focus on developers on BitCoin and block-chain technologies will help keep the security situation up to date.  I think security through transparency is the best policy and the open source nature of BitCoin and the momentum of developers around it is a good thing.

Paul Brody
Paul Brody

Certainly written by an IBMer - me.  To the extent that it represents our views, then yes, it's "marketing" for IBM.  That being said, I personally strongly believe in this approach.  I like the BlockChain because I think it can address some very critical business issues for our clients while also being a foundation for a more secure and re-decentralized Internet.

andypiper
andypiper

@theRab I used to be big in the Internet of Things. I knew some stuff.

jg_networks
jg_networks

@S_dF maybe it's the catalyst we need, but in the meantime [horror intensifies]

Gene Vayngrib
Gene Vayngrib

@Paul Brody  It was interesting to learn that IBM is embracing the blockchain. Blockchain shapes up to initiate a new architecture shift, at the level at which cloud was, or possibly bigger, at the level of change which Web introduced in 1994. Paul, I see that you have supply chain background. My team has spent over a decade in supply chain management trenches applying decentralized approach to coordinating the activities of participants in 20+ countries. We are now creating blueprints for doing the same but much better based on the blockchain. May be we can chat. I am writing a whitepaper on that subject github.com/urbien/tradle/blob/master/README.md 

gigastacey
gigastacey

@pbrody Come talk about it on my IoT podcast. The Telehash stuff was intriguing.