Subscribe to DSC Newsletter

Q&A: Andy Stanford-Clark on IoT [Part 1]

As organizations look for strategic direction on the Internet of Things (IoT), it’s their developers who are likely to be the strongest advocates for sifting out which technologies and standards can provide the best competitive edge.

Andy Stanford-Clark (@andysc) is the Chief Technologist for Smarter Energy in IBM’s consulting business in energy and utilities for IBM U.K.  He’s also an IBM distinguished engineer and a master inventor, “Which means, I’ve got quite a lot of patents, and I invent cool things for IBM,” he quipped. Stanford-Clark is also known for creating the IoT connectivity protocol, MQTT.

Stanford-Clark is scheduled to keynote at the EclipseCon Europe 2014 event where he will discuss the importance of the Internet of Things to the development community. Developers from about 30 countries are expected to attend the event. Most work for companies and enterprises that build commercial products or internal tools based on the Eclipse framework, such as BIRT.

Actuate got a chance to catch up with Stanford-Clark and chat about the Internet of Things and where developers should focus their efforts.

Actuate:               First off, what’s your general overview of the Internet of Things, and what kind of impact that it is having on the businesses that you’re working with?

Stanford-Clark: So, I’ve actually been working in what we now call the Internet of Things for about 15 years.  Back in 1998, I moved back to the U.K. from three years in the U.S., and started working on remote telemetry systems for oil and gas pipelines for SCADA (supervisory control and data acquisition) systems.  And it quickly became apparent that we needed to create something which brought that industry into what was then the modern world, with TCP/IP, and packet-switched networks and so on.

So, I developed a protocol called MQTT.  But that was something I developed in 1998.  And the good news is, this year, it’s about to become OASIS standard.  So, finally, it’s made it through all the hurdles and the amount of market penetration and so on to get to a state where it’s standardized.  That’s very exciting for me.

But, along the way, over 15 years, we’ve called it many things: remote telemetry, pervasive computing, ubiquitous computing (ubicomp), tier-zero … even M2M [machine-to-machine] is still around.  IBM came up with the term Smarter Planet about six years ago, where we talk about instrumented, interconnected, intelligent devices in the world.  And that was effectively our name back then for what is now the Internet of Things.  That term has taken over, and it’s become the sort of lingua franca for all that stuff.

So, largely it’s the same things as we’ve been doing all along, but a shiny new name for it.  And the Internet of Things is the current buzzword du jour for that technology.  So, nothing new here, but in terms of the penetration, and people’s understanding of it, and the divergence of products, and services, and thinking, and ideas, and companies all over the world, wherever you look, that’s really exciting.

And I think it’s really snowballed into something which is kind of on everyone’s lips.  Everyone’s thinking about it, and people are understanding where it fits into their world, and what the benefits of it are, and how you can integrate data from lots of different senses to provide better services for the public, or for your clients, or whatever it is.  So, that kind of sets the context.

Actuate:               So, where do you feel the tipping point has come for everybody to now acknowledge IoT as lingua franca? Do you have richer conversations now that people are getting their heads around what Internet of Things might mean to their organizations?

Stanford-Clark: Yeah.  So, I think the tipping point has really been the commoditization of technology.  So, when I started this stuff for real, I started doing my own home automation and home energy monitoring stuff, which was about 10 years ago.  I was kind of soldering all the circuit boards, and, you know, inventing the sensors, and linking it all together by hand, whereas, now, you can buy off-the-shelf, very cheaply, a home energy monitor, and link it up to the Internet, and get the data from it in a nice JSON feed, or MQTT data, or whatever it is.

And so that really becomes the tipping point.  So, we moved from the domain of the rich geeky guy to anybody to be able to go out and buy this stuff, and start to integrate stuff into their homes without needing a PhD in computer science or electronics to get it all working together.  So, that’s really been the tipping point, from my point of view.

In terms of richness of conversation, first of all, people are starting to look at applications in a broader context than what they did in a traditional client-server or web e-business. They are starting to think about an agenda around mobile computing, social media, the analytics that wrap around that, the idea of hosting services in the cloud – as part of the system is kind of out there in the field, and the other part is up in the cloud.

[Because of cloud computing], you just have this huge ubiquitous connectivity to get to it.  That broadens the scope for people, thinking about how they interact with their customers, their clients, their employees, the rest of the world.  And people are starting to think about, “What is my Internet of Things agenda?”  “What’s the strategy for the next few years?”

So, rather than building, say, an app to access your banking system, you start to think in much broader context about, how does the Internet of Things include mobile and start to become integrated into the whole business context.  And to my thinking, that becomes real, when a real-world event becomes a business event in the enterprise.  So, something kicks off a work process, or a transaction; something in your enterprise messaging system, which is caused by some event picked up from some sensor, or an RFID tag read, or a person’s phone entering a geo fence, or whatever it is, out in the real world.

Actuate:               Your expertise has been within the energy systems, and utilities, and such.  Anything specific there that’s made a shift now that people are connecting the dots for the Internet of Things?

 Stanford-Clark: Yeah, very much so.  Now I see Smart Energy as being sort of an application of the Internet of Things.  So, I happen to have specialized in the energy sector after the last few years.  That in itself is a sort of subset of Smarter Cities, which then is part of the Smarter Planet kind of agenda.  So, it kind of fits nicely.  It’s like Russian dolls inside each other.  It’s all Internet of Things, really.

And the interesting thing from my point of view there is that the smart grid that we talk about isn’t one thing; it’s a whole load of things, a really heterogeneous mix of sensors, actuators, people, things, components from different manufacturers like solar panels, wind turbines, electric vehicles, grid storage mechanisms.  All that stuff, in order to operate holistically as a grid system, all needs to be instrumented, interconnected, and then use intelligent algorithms to model the activity and behavior of those things, bringing in information like weather forecasts, and demand forecasts, and so on, and run the whole thing as one big holistic system. Which, of course, means you’re going to have sensors on everything that’s then all connected into other central systems, or, more likely, federated hubs of information to make intelligent decisions about how to operate components on the grid in order to make the whole thing work as one big happy system.

To my mind, the smart grid is a classic example of how the Internet of Things needs to come together across this really heterogeneous mix of different manufacturers making different devices, in order to make an ecosystem operate holistically.

 Actuate:               Can you provide a concrete example of how IoT is succeeding in this today?  Do you have a personal favorite, something maybe you’ve worked on, or something that a partner might be doing?

Stanford-Clark: In terms of IoT at scale, I don’t think I have any good examples, but one area very close to my heart is home energy monitoring and some of the sort of spinoff things you can do within the home to get further insight.  So, for example, if you’re monitoring the appliances in my house, you might notice that the TV’s on or the kettle’s been on to make a cup of tea or something, and that’s kind of interesting, but not particularly valuable.

But say I’m monitoring my aging parents, and their system’s wired up as well in their house, and at 10:00, I get a text saying, “Hey, your mom and dad didn’t have a cup of tea yet today; maybe you should give them a call.”  So, you start to get into this sort of remote monitoring of elderly or sick people who have difficulty coping with normal life, being able to, not monitor them electronically, but, to get a head’s up when something’s maybe amiss, so that you can intervene more rapidly.

I think there’s huge potential there, in terms of if a window’s been left open, maybe if the front door opens in the middle of the night; maybe something bad is happening.  Just to be aware of those things happening.  If you’re away from home; say you’re monitoring your holiday home, if you have one of those, to just find out what’s going on, to build that picture of information derived from the data.  And it might be completely innocent, or it might be something that you actually need to know about.  And rather than getting deluged with tons and tons of information, little data points, you actually get meaningful notification.

We term it actionable insights.  So, this means you take data, transform it into information and knowledge about what’s happening.  And then if something important happens that you should know about, we create an actionable insight, which is maybe a text alert, or a tweet, or a red flashing light in my room, or maybe a notification to my car dashboard, something which tells me there’s something happening that I should know about.  And that, to me, is a really exciting, sort of multidimensional data opportunity for the Internet of Things which is really just emerging at the moment.

Actuate:               If the Internet of Things is fueling even more of these actionable items, what part do you feel that visualizations and dashboards can play in the success of those IoT developments?

Stanford-Clark: So, what I’m seeing is, the dashboards and visualizations are extremely important to understand the data in the first place.  Remember, I talked about the three “I”s of instrumented, interconnected, and intelligent.  The first step in intelligence is to know what’s going on.  And we found visualization in various forms, from traditional graphs, through ambient devices – a glowing orb device in my house that shows me red, amber, or green, how much power we’re using – through to things like tweets, which also are a form of visualization of information in textual form, through to banks of multicolored LEDs which have a picture of data rippling across them to give you real-time information.

All that stuff gives you an understanding of what’s happening.  And I found it very valuable to then develop the algorithms so the computer can sit watch those dashboards and those lights for me, to give me those actionable insights, to send me the notifications.  And so it’s just kind of the first step in understanding the data, to look at it, visualize it, graph it up, and see what’s happening.  And then you can say, “Ah, okay.  We’re seeing that kind of spike, or that kind of ramp, that kind of event happening.  I can see how to program the computer, to define some rules, to look for that kind of thing.”

In a wider context, being able to see what’s happening is extremely important, and converting the data into pictures.  I use the horrible cliché – a picture’s worth a thousand words – but it really, really is.  And watching the visual information is far more likely to show the anomalies in the information that you’re trying to see, if it’s in some kind of visual form.


This conversation will continue in Part 2.

- See more at:

Views: 347


You need to be a member of Data Science Central to add comments!

Join Data Science Central

© 2021   TechTarget, Inc.   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service