And another thing 66

by Jon Honeyball

Jon Honeyball gets wound up about the Internet of Things.

HardCopy Issue: 66 | Published: June 1, 2015

Some people are getting very wound up about this Internet of Things thing. How it is going to be transformative, how it will change our lives, how things will never be the same again. Frankly, I could say the same about almost any piece of technology that I have bought over the last 30 years. When I look back and see just how primitive things really used to be, I am actually quite staggered by both the rate of change, and the actual lack of change.

Now I have to explain that seemingly impossible contradiction.

Network bandwidth

Are we sleepwalking into a corporate network bandwidth crisis? I am not just talking about the connectivity you need to the outside world, or connections between sites – the connections within a site are critical too.

Last week, I attended the National Association of Broadcasters conference in Las Vegas, a huge show where everything to do with professional audio and video happens. Here it became apparent that suddenly, the world has woken up to 4K video. Not just highly compressed 4K like that which Netflix dishes up, but the real thing running uncompressed in its 4:2:2 60p framerate loveliness.

The data rates involved are truly staggering; at over 18 gigabits per second for one UHD60p uncompressed data stream, you have to start thinking in entirely different terms when it comes to moving this sort of information around. There are a few companies I keep an eye on for this, such as Black Magic Design. In one fell swoop it has upgraded most of its TV, broadcast and transmission/distribution technology to ‘12G’, which is the Full Monty of UHD60p. This includes such behemoths as a 40-in 40-out cross point matrix running full speed. The data rates are enough to make your hair curl.
So why does this matter to the IT world? Because at some point it is going to land on an IP network. You might hope that by then it has been massively compressed, squeezed and knocked into a manageable shape, but one thing is clear; broadcast quality video is making a gigantic leap in data rates, and it will impact corporate networks.

Even today, we have limited real implementation of proper QOS systems outside of the most rigorously run and highly managed networks. VLANS are found, but there is a mindset that says that 1Gb Ethernet is all you need for any possible use in the future. I am far from convinced that this is true moving forward. What is clear is that these things tend to creep up on you when you are least expecting it, and maybe it is time to start thinking about how we will deliver 10G to the desktop. Of course, an Excel worksheet doesn’t have that load requirement, but video is one area where you can never have enough bandwidth. And it would be maddening if your internet network gets left behind, because when video comes to party, it is the only elephant in the room.

Back in 1990 I bought Word for Windows 1.0. It actually ran on Windows/286 and we had to wait a few months for Word for Windows 1.1 to come along and support Windows 3.0 with its nicely chiselled grey buttons. Now here I am, some 25 years later, typing this text into the beta of Word for Windows 2015 (or is it 2016?) for Mac. Of course this new version has loads of new things to play with, but the simple truth is that there isn’t anything like 25 years of progress here – not 25 years’ worth of work by a well-staffed development team.

It’s all different on the hardware front, of course. Now I have boundless CPU power, coupled to GPU loveliness, running an operating system on hardware which has an incredible amount of increasing powerful components, and battery life of well over 12 hours. The superb performance given to us consistently by Moore’s Law over the past 40 years cannot be overlooked.

So in our core computing, I am far from convinced that we have experienced this generational shift which others claim. And let’s not get all romantic about web-based application design: after all we did have client-server activities before that.

OK, the arrival of the smartphone and the tablet cannot be overlooked, but I am not sure that these can really be counted as true IoT devices. For sure, they connect up through technologies like Bluetooth and let me control various things between them. The arrival of smart watches and embedded devices is another step towards IoT, but for the moment they remain far too tethered to a companion device for me to be convinced they have a real IoT role to play. Why should I need to own an iPhone when I have an Apple Watch? Why should my Microsoft Band require a Bluetooth connection to software running on a Windows Phone? Why can’t my Samsung smart watch talk directly to my Samsung TV, cutting out the mobile phone component completely?

Apple Watch and iPhone

Very nice, but why do I need them both?

Until we have a situation where any device can talk to any other device, irrespective of size, I am not sure that we can claim to have a true Internet of Things. To make this really work we need a ‘lingua franca’ that will allow apps running on disparate devices to communicate, wherever they are running. It certainly shouldn’t require private APIs and formats that allow specific private conversations to happen. And where is auto-discovery in all of this? Another pipe-dream that we have to wait for yet again?

No, all I can see for the time being is a point-to-point connected mishmash, with no real thought as to how we will actually use these things in day-to-day life. If you fancy a challenge, try sorting out the problem I had to solve yesterday. On my motorbike, I have a Garmin GPS unit. I have a SENA Bluetooth headset mounted on my helmet, and an iPhone 6 Plus in my pocket. I want to connect the iPhone to the Garmin, so I can run the Garmin Connect+ app which feeds real-time traffic data to the GPS. I want to connect my iPhone 6 Plus to my headset so I can send and receive calls, all via the rather useful SIRI voice controlled interface. And I would like to connect my Garmin to the SENA headset so I can hear traffic and directional routing information as I am riding along. Trying to get this menage-a-trois to actually play nicely is nearly impossible. Add in a wrist mounted watchband device that will vibrate when something needs my attention, and the complexity goes out of all semblance of control. If we were truly in an IoT world, then these things would just work. But they don’t, and I cannot foresee a time when they will.

Of course, I understand why it is desirable to have the Apple Watch chattering away to an iPhone, and that it’s not likely to party with an Android device any time soon. I understand how it makes the programming interfaces very much easier when you have the Apple Phone style of communications protocol stack working seamlessly over Bluetooth. But even Microsoft, that arch enemy of cross platform development in the past, has woken up to the new world order. My new Microsoft Band connected up just fine to the Microsoft app running on my iPhone. How the mighty have fallen.

Not getting these things working together in a seamless, truly connected IoT way is what will kill the whole premise of the IoT platform, unless these problems get solved soon. And that means someone has to stick their neck out and really want to make IoT interoperation work.