Comments
Internet Of Things: What's Holding Us Back
Threaded  |  Newest First  |  Oldest First
Laurianne
50%
50%
Laurianne,
User Rank: Author
5/5/2014 | 10:57:28 AM
IoT software vs. hardware
Tennison's story about asking analytics software vendors to solve his sensor hardware problem is striking. It makes me wonder if we are much further along on the software side than the hardware side regarding IoT.

Also his problem is specialized to the railroad industry -- and there will be examples that need to be customized for every vertical. How will sensor manufacturers achieve affordable scale? Lots of interesting food for thought here.
ChrisMurphy
50%
50%
ChrisMurphy,
User Rank: Author
5/5/2014 | 11:34:27 AM
Re: IoT software vs. hardware
The lack of deep concern about analytical capabilities was one of the biggest surprises to me from the reporting in this article. If others have had a different experience, glad to hear about it.

In terms of sensors, we'll see about how much industry-specific adaptation is needed. Bill Ruh of GE noted that in the mechanical world, measuring vibration is something of a universal need -- it's rarely a good thing with machines. However, not all the innovation in sensors will be hardware driven -- virtual sensors, or software-based sensors that combine inputs from multiple sensors, will emerge as companies needs get more sophisticated.   

 
Drew Conry-Murray
50%
50%
Drew Conry-Murray,
User Rank: Ninja
5/5/2014 | 3:51:42 PM
Re: IoT software vs. hardware
I'm surprised about the lack of concern about analytical capabilities too. The security industry has been wrestling for years about how to identify actionable information from massive quantities of data. Maybe with IoT you're dealing with a small subset of known data types (pressure, temperature, vibration, etc), which makes it less about needles and haystacks.
Todder
50%
50%
Todder,
User Rank: Strategist
5/5/2014 | 4:34:00 PM
Re: IoT software vs. hardware
The railway and oil industry requirements are the same for the most part (like any other industry, resource-based or other that are remote from urbanania).

Remote locations, little WiFi, power in short supply, and the environmental challanges. Solar powered units would resolve some of this potentially but you also need some fail safes built in to any IoT application. Like when the power to traffic lights goes off at an intersection, you can get a Bobby to manually direct traffic. Any implementation requires an audit trail of real-time testing that it is working and a backup (tested to fall over to when it isn't).
Thomas Claburn
50%
50%
Thomas Claburn,
User Rank: Author
5/5/2014 | 4:51:32 PM
Re: IoT software vs. hardware
I find it interesting that the IoT is hobbled by the same thing that limits smartphones: power. We need a breakthrough in power storage and generation that improves current technology by an order of magnitude, something along the lines of Iron Man's arc reactor.
ChrisMurphy
50%
50%
ChrisMurphy,
User Rank: Author
5/5/2014 | 4:57:11 PM
Re: IoT software vs. hardware
The power limitation is definitely true, Tom. One difference from smartphones is that industrial uses often involve relatively low power demands -- sending small bits of data back, but needing to stay powered over many months because replacing/recharging is difficult.  
Brian.Dean
50%
50%
Brian.Dean,
User Rank: Ninja
5/6/2014 | 2:17:20 AM
Re: IoT software vs. hardware
Without electricity, communications and the internet could not exist, without the internet, Cloud computing would not exist, I think that the majority of the IoT will need the Cloud. Hence, without the Cloud a lot of IoT's potential might not materialize. The inter-linked and dependent environment of technology might provide a break-through, but I feel that this break-through might just be written off as the normal pace of technological progress.
SaneIT
50%
50%
SaneIT,
User Rank: Ninja
5/7/2014 | 7:32:40 AM
Re: IoT software vs. hardware
The really large IoT solutions will need the cloud to be truly effective but right now those solutions are few and far between.  I can't think of many companies that cover the area that Union Pacific covers with the need to monitor many points along the way.  Even companies like UPS or Fedex don't need to see the condition of every mile of road that they travel.  I think some government projects could become this big but if you're looking to track trucks for a medium sized company via GPS then moving to the cloud isn't going to be as critical for success.  
WKash
50%
50%
WKash,
User Rank: Author
5/6/2014 | 2:53:51 PM
Re: IoT software vs. hardware
Insightful article, Chris. Unless you've been out to the remote areas where these pipelines and railines have been built, it's easy to underestimate just how far-flung and distant this infrastructure is -- and what it will take to develop sensors that would stay powered and in touch. But I sense the integration of data question may be equally challenging, and another reason why the IoT will work in some fields, and take much longer in others.
Laurianne
50%
50%
Laurianne,
User Rank: Author
5/5/2014 | 5:03:32 PM
Re: IoT software vs. hardware
We need a bandwidth breakthrough, also. Bonus points for connecting Iron Man to this debate.
DanC364
50%
50%
DanC364,
User Rank: Apprentice
5/7/2014 | 4:28:16 PM
Re: IoT software vs. hardware
At Packet Dynamics we realized that the 'dark territory' or areas not covered by cellular M2M services exist not only in the U.S but also globally. Infrastructure players whether rail, pipeline, electrical transmission, etc. all need to communicate to end nodes that might never be covered by GSM/LTE. The Cloud can't just exist in the colored areas on AT&T, Verizon, and Sprint's coverage maps. 

We provide beyond-line-of-sight communications for M2M and cloud applications using High Frequency radio links.  We went back to HF, an old school spectrum that still works, and added the latest in radio technology (waveforms, error correction, etc) to solve real world industrial problems. It's not the glamorous side of the Cloud, but for companies with equipment scattered across the fly-over states or in third world countries it does solve a major problem.

Dan Conner

www.Packet-Dynamics.com
ChrisMurphy
50%
50%
ChrisMurphy,
User Rank: Author
5/9/2014 | 4:08:07 PM
Who Gets the Data?
An interesting idea that I didn't explore, that I heard in a follow-up chat this week with the folks at M2Mi (thanks Manu), was about who gets access to the Internet of things data. Jim Fowler, quoted in the article, talked about this at the IW Conference -- aobut how GE only gets access to power plant data if the customer continues to see the value in sharing it. But the M2Mi folks brought up other parties who will want access -- for example, the lender who owns a piece of equipment you're leasing, to make sure you're doing the maintenance so it'll have the expected residual value at lease end. A lot to think abot there in terms of who gets access to the data beyond the simple Maker-Owner.  
mnamboodiri
50%
50%
mnamboodiri,
User Rank: Apprentice
5/5/2014 | 1:49:29 PM
IoT with more capable platforms
Chris - interesting article. I think the way we will be able to address some of this complexity and integration challenges is with the platform (or middleware) taking up a larger burden. If we continue to custom build each solution (as the article describes we do today) with massive integration efforts, security, burdened devices and heavy apps that have to embed connectivity, networking, data massaging, QoS etc, it will be a long time before we get to 50 billion useful connected devices! 

I believe some of the burden will be delegated to platforms that can then enable more focused devices and "thin" apps - while providing the networking, security, contextual intelligencem modularity, data flow/access and APIs to build solutions faster. 

-Manu
ChrisMurphy
50%
50%
ChrisMurphy,
User Rank: Author
5/5/2014 | 2:17:46 PM
Re: IoT with more capable platforms
Manu, thanks for the perspective. It does seem like there's a role for a management platform in here somewhere that is still evolving today. 
SteveJ447
50%
50%
SteveJ447,
User Rank: Apprentice
5/5/2014 | 5:20:41 PM
Delivering the data
Interesting that the focus is mostly on the edge (e.g. sensors) and the centre (e.g. cloud-based analytics) without much discussion of the data-sharing infrastructure needed to support any end-to-end system.

Whether the system designer leans-toward sensor-based (edge) computation or, alternatively, 'thin' apps connected via a cloud service, it is inevitable that both will be required in any business-critical system, and furthermore, device-to-device data-sharing (e.g. for local real-time control) will often be required too (e.g. when the latencies to/from the cloud are too long). So a lot more attention needs to be given to the real-time data-sharing platforms that will be needed to underpin and enable these IoT systems. Many system designers will recognize this as the 'elephant in the room' since commentators tend to focus on: 1/ smart sensors and 2/ big-data analytics, while assuming some kind of wireless connectively alone is sufficient for real-time data-sharing. Fortunately the technology exists (e.g. the OMG's DDS specification) and is standardized and proven. Hopefully we'll start to see a lot more discussion of the system infrastructure...rather than just the data sources and sinks that it connects. Ubiquitous, real-time, secure data availability won't just happen.              
Joe Stanganelli
50%
50%
Joe Stanganelli,
User Rank: Ninja
5/10/2014 | 6:42:32 AM
Re: Delivering the data
And then there's the matter of encasing the stuff in lead/protecting it against solar flares, seeing as how we're rather a bit due for another Carrington Event.
Gary_EL
50%
50%
Gary_EL,
User Rank: Ninja
5/10/2014 | 2:37:08 PM
Re: Delivering the data
It'll be much worse this time. 19th Century telegraphs were able to go back to work after the event was over. 21st century CMOS will be PERMANENTLY RUINED. The machines that will be needed to build replacements will be PERMANENTLY RUINED. Much of the specs, stored in semiconductor or magnetic media will be permanently lost, as will be any electronic equipment built after the late 1960's.
Joe Stanganelli
50%
50%
Joe Stanganelli,
User Rank: Ninja
5/11/2014 | 4:42:33 AM
Re: Delivering the data
I remember reading something about a bill that would protect a lot of transformers and other vital parts of the nation's infrastructure...but it hasn't passed.

Ridiculousness.
jayfriedmn
50%
50%
jayfriedmn,
User Rank: Apprentice
5/5/2014 | 7:21:35 PM
FedEx for Data
Good perspectives.  IoT, especially in the industrial world is not as simple as a NEST appliance.  There are several considerations here:  the device, the data, the analytics, and then the data tranport mechanism.  On top of that, there are almost 2 billion legacy devices already spitting out data but until now there hasn't been an easy way to secure and backhaul that data.

The data transportation question, is equivalent to FedEx for packages.  An organization just wants to securely transport data from point A to point B.  You don't really care how the package (or data) gets there, just that it does on time.  And, most of todays sensors are also still serial devices - not IP-based.  

This is where virtual networks through software-defined architectures meets the Internet of things.  Virtualization makes the problem a whole lot easier to solve.  

 

Jay Friedman

Distrix (www.distrix.com)
danielcawrey
50%
50%
danielcawrey,
User Rank: Ninja
5/11/2014 | 1:26:31 PM
Re: FedEx for Data
Interesting article. I don't think that many of us are thinking about the cost of resources for IoT. This includes the cost of bandwidth and power as well as implementation. Right now, based on the quality of data that is coming out of many devices, it's simply not worth all of the initial and operating costs. 

We've been hearing about the Internet of Things for years. But it seems as if this is going to take much longer to be a reality than many prognosticators have believed. 
Joe Stanganelli
50%
50%
Joe Stanganelli,
User Rank: Ninja
5/10/2014 | 6:39:38 AM
Getting Off Track
I love how this much technological planning has to go into trains -- which FOLLOW A TRACK -- yet Google and others keep pushing the idea that the consumer market is totally ready for self-driving cars.
tjgoodmanson
100%
0%
tjgoodmanson,
User Rank: Apprentice
5/13/2014 | 2:34:24 PM
Directional Analytics = Speed
We all know the amount of data we collect continues to expand at unprecedented rates, and I agree one of the main barriers to this data is the ability to analyze it efficiently. Data analysis does take expertise, but I believe the onus to employ data scientists should be on the software and analytics vendors — not the end users. 

If we make software that provides directional analytics from all of these things, creating an "if this, then that scenario," more people will be able to benefit from that data and act on it sooner: rather than waiting days, weeks or months for direction from an analyst. 

And when we're looking at examples of railroad switches, oil wells and even household appliances: time is money. 

The beauty of big data is we have access to historical data — from single devices to entire infrastructures — as well as real-time live data. And if information from a sensor had triggered a certain event or scenario in the past, the software measuring that sensor can rely on that historic data to recreate (or prevent) that scenario again. 

 

Tom Goodmanson

www.calabrio.com

 


The Business of Going Digital
The Business of Going Digital
Digital business isn't about changing code; it's about changing what legacy sales, distribution, customer service, and product groups do in the new digital age. It's about bringing big data analytics, mobile, social, marketing automation, cloud computing, and the app economy together to launch new products and services. We're seeing new titles in this digital revolution, new responsibilities, new business models, and major shifts in technology spending.
Register for InformationWeek Newsletters
White Papers
Current Issue
InformationWeek Tech Digest - July 22, 2014
Sophisticated attacks demand real-time risk management and continuous monitoring. Here's how federal agencies are meeting that challenge.
Flash Poll
Video
Slideshows
Twitter Feed
InformationWeek Radio
Live Streaming Video
Everything You've Been Told About Mobility Is Wrong
Attend this video symposium with Sean Wisdom, Global Director of Mobility Solutions, and learn about how you can harness powerful new products to mobilize your business potential.