Things I Find Interesting
Who’s Going to Drive IoT Innovation
Nov 20th
Let’s face it, the Internet of Things (IoT) is hot right now. And it’s only going to get hotter. Everyone wants in on the perceived coming gold rush that will be the Internet of Things. And a gold rush is most definitely predicted! Gartner has said so here, and here. IDC says so , too. I could go on and list how many analysts and media pundits, etc. agree, but that would just be redundant. Suffice to say that basically everyone thinks so. And they’re all probably right.
To some of us, this isn’t a surprise. For instance, back in 2004 I started saying the following:
“Sun has licensed Java™ on a billion cell phones. Now, if every person on the planet carried 3 cell phones, you’re still only talking about maybe 15 billion devices. That’s a limited market. But if we were to license Java™ on every processor in your car, your kitchen, your building, etc. then you’re talking trillions of devices. Now we’re talking about a real market!”
Yes, I said that. In 2004. And that was before there was any “Internet of Things” in any real sense. I’ve been repeating it ever since. But we (at Sun Labs) were developing a sensor platform for the internet of things, and that’s the vision we had. Score one for us.
We also saw that the cost-curve on 32-bit processors would bend significantly and that in 10 years, embedded devices would all be based on 32-bit, powerful, cheap processors, so that’s what we built back then. That was the Project Sun SPOT.
Fast forward 10 years. Here we are, with a plethora of devices from Raspberry Pi, to Intel Edison to Electric Imp to Spark, to any of the hundreds of Arduino-based devices. 32-bit sensor platforms are the default, not the outliers. Score one for us. And as noted above, all the analysts are predicting explosive growth in the IoT market space. Score two for us.
One of the 3rd assumptions we made was that Sun SPOTs should be a “platform for innovation” and that Makers, the guys in garages and basements who do this in their spare time, because they love it and can’t stop (guys like me, frankly) would be the market movers. We ‘marketed’ to the Maker movement. We had a booth at the very first Maker Faire in San Mateo, CA. Go see the kinds of crazy things the kids were doing with Sun SPOTs back in the day. The funnest projects were done by regular Makers. Guess what? Score 3 for us. See?
Yes, the big guys like Intel and ARM will move the hardware needle. They will be pushed by the Arduino folks, in many ways, but most basement/garage makers aren’t going to be developing processors, though they may (and probably will) develop boards and platforms based on processors by ARM and Intel and others. Supported by places like SparkFun (I love them, and have been a SparkFun junkie since ’04!) Makers will develop sensor platforms and implementations that will move the market. And what will be key to the Makers success? Aside from the availability of great platforms, the great enabler will be Open Source. Open Source Hardware and Open Source software. Open APIs, open software for authorization and authentication and configuration and management. No Maker can afford to buy an enterprise license for your LDAP/Authentication platform for their project. No Maker will want to pay a license fee — at least initially — to use your APIs. They want to build a neat project. They want to solve a specific problem. Once they succeed with that project they may turn it into a product. That’s when the API company, the Authentication company, etc. can expect to get paid.
So if you’re one of those companies. If you are selling APIs, or authentication and management, or whatever it is, to the IoT market, you’d better make it Open Source. You’d better allow the Makers of the world to download it and use it for free to build their project. Because they will be the next Google. The next Facebook. The next Instagram. And when they are, you’ll want to be the one that was their (free) friend back in the day.
Call me crazy. Tell me I don’t know what I’m talking about. But I refer you to the above. 3 for 3. I’m not saying I’m a genius, or that I’m Carnak the Magnificent. I’m just sayin’.
Drop a Brick On It
Nov 19th
I’m all about connectivity. Yes, there are times to disconnect from the world, but that’s a personal disconnect. Applications and devices should, for the most part, be connected all the time. But connectivity is expensive and hard to maintain in remote places. If you’re connecting a server, or your house, there are many options available — let’s just go with that for now, shall we? The whole ISP monopoly issue will just get me wound up. Typically these kinds of connections are fast, reliable, and stable. Mostly they’re cheap and widely available as well.
But when we look at the Internet of Things (ioT) things can get a bit more squirrelly. Yes, if your IoT devices are things in your home, you can go with the connection you already have. If they are enterprise-based IoT assets, you have your enterprise-class connectivity to rely on. What about IoT devices that are either field-based or mobile? What if your IoT deployment is in a jungle in Panama? Or the dessert in Africa?
I found this the other day, and it seems like the perfect solution. Drop a Brck on it! This looks like a great device for remote connectivity, especially for IoT or low-bandwidth applications. There are a couple of really cool looking features here — and a few that I’d like to investigate further, if I had the time and the device. The 8-hour built-in battery backup for this device is a huge win in many remote deployment scenarios, to be sure. The fact that this device was developed in Africa specifically for remote deployments is also a huge plus. Again though, there are a few things I think need further investigation.
First is the ability to add Arduino-Based sensors and devices directly to the box. So if you’re using Zigbee or IEEE 802.15.4 devices as your end-nodes, you could add a gateway to those devices to the Brck devices directly. If you’re using BTLE devices, you could add a BTLE-gateway for access to those devices as well. This was — back in the day — one of the big sticking points for lots of sensor applications. Low-power, low-energy network connections to the devices were hard to bridge to the internet for backhauling the data. In many deployments, it still is. Especially for remote-sensing applications. Being able to build-in your internet-to-sensor gateway device would be a huge plus.
What I would love to investigate further — given the time and the device, of course, since IoT is my hobby, not my job — is whether or not I could deploy actual services to the Brck device. Like authentication, authorization, data integrity and IoT device management. I know, from their website, that they have cloud-based tools — who doesn’t, amirite? — to manage your Brck devices, and they say you can push arduino sketches down to the box to distribute to connected devices, but what about these other services? What if I wanted to use, say, OpenAM and OpenIDM services? Could I deploy these services directly to the node? Or could I build the service calls into the device to proxy for the end-nodes?
I guess I’ll just add this to the list of devices and technologies I’d love to be able to spend some time with, but probably won’t have the time or the money to deal with. <sigh>
Making a Button
Nov 18th
I can’t resist. I mean really. Who doesn’t need a big red button that you can program to … act like a big red button? But it’s so much more than just a button! I mean, a button is just something you push, right? This? This is so much more!
The bttn communicates with the powerful bt.tn servers in the cloud. They dispatch messages and actions to almost any internet services so that what you want gets done. When the Magic has happened the bt.tn servers notify your bttn that the deed is done.
Because what every Big Red Button needs is powerful servers in the cloud standing behind it! Makes me want to make one with a “Panic” sign wonder it. Or a big sign under it that says “Do NOT Press!”
In all seriousness, I can think of some actual practical uses for this thing — assistive device for my son, a big red doorbell, etc. — but none that really rise to the level of requiring me to actually purchase one.
Doesn’t mean I don’t want to though!
Feel free to send me one though. http://bt.tn/
Hardware Junkie
Nov 13th
Ok, so in addition to being a gadget freak, I’m a hardware junkie. It’s been a while since I’ve actually done much hardware hackery, but that’s not because of a lack of desire.
I bought an RFDuino dev kit last year as a Kickstarter backer that I just opened last weekend, and I haven’t done much with it so far. Turns out having nothing but BTLE is a lot less useful than I thought.
I have almost 1,000 Sun SPOT devices squirreled away here in the house for ‘future use’ — unless you want to buy some of them, and if so, let’s talk!
But this device — the Spark Photon — this device looks irresistible. And not just because of it’s cute size either. It’s packed with cool features, and the development environment looks to be really nice, too! As they say on their website, “When you write embedded code, you can choose between Wiring (the language of Arduino), C/C++ or even ARM assembly, depending on your level of experience and self-hate.” I don’t have enough self-hate to want to write in assembly.
Spark’s hardware development kit, the Photon, provides everything you need to build a connected product. We combine a powerful ARM Cortex M3 micro-controller with a Broadcom Wi-Fi chip to get you started quickly. The design is open source, so when you’re ready to integrate the Photon into your product, you can.
So there’s that. But there’s a lot more that looks attractive other than the open source aspects. I’m especially intrigued by the built-in REST APIs right now, for a variety of reasons. Being able to call both into and out of the device via REST opens up a whole bunch of interesting possibilities with them. They could be command and control modules for the Riverbed SteelHead, SteelFusion and SteelCentral appliances. They could easily be integrated into an Access Management System like the ForgeRock OpenAM one.
In addition, it appears — at least from the documentation — that there’s security built-in with RSA, AES and SSL/TLS out of the box. Security, built in from day zero, is key to a reasonable IoT device as simply attempting to add security as a layer later on is a recipe for disaster — and insecurity/vulnerability.
I have boxes and boxes and boxes of sensors and other small gadgets I’d love to try hooking up to one of these devices. The only trouble is, I’ll have to wait until at least 10,000 units are pre-ordered before they actually manufacture them. So get ordering!!! I’m Pre-ordering now!
I’m a Gadget Freak
Nov 7th
I freely admit that I am a gadget freak. I’ll also (sometimes) admit that it’s a problem. I bought a Cookoo Watch as a Kickstarter Backer of the project. I own a FitBit — I’m on my third one, don’t ask. I’ll probably buy an Apple Watch. But one of the problems I have with most ‘wearable tech’ is that, well, it looks too much like tech.
The Cookoo Watch was a great start, but the first version was really buggy and unstable — except as a watch — and I’ve come to find out that it’s un-upgradeable and is basically a paperweight. So much for being an early adopter there.
Going out on the town? The FitBit has to come off, because, well, it is a plastic thing that looks tacky. So no activity tracking while you’re out. Plus, wearing a FitBit and a watch clutters the wrist. So what to do?
A few months ago, when I first started wearing the FitBit, I went looking for any alternatives that were’t so ugly. I found the Withings Activité Watch. It does all the things a FitBit does — more actually — and it doesn’t look like Tech. It actually looks like a watch! And an analog watch at that. I don’t like digital watches. For some reason, I can tell time on a digital watch. I need to see the hands on a clock face to tell time. Yeah, I’m weird like that.
It comes with the HealthMate App, which looks a little more extensive than the FitBit App. The HealthMate App will integrate with the Apple Health App which, up to now, is pretty useless.
So this watch is a start, but it’s still not everything I would like in a wearable tech device. What would I ultimately like? I’d like a fully functioning watch, like the Withings that is a real watch, that has a sleep tracker, a steps tracker, and all of the health-tracking niceties. I’d also love it if it would either do a heart-rate monitoring, for when I’m running or exercising, or connect to my Heart Rate band that I wear with my Suunto watch while I work out. In essence, I’d like to have just one device. Not one for working out, one for steps, one for … you know, one that does it all.
New Blog!
Oct 23rd
As the year draws to a close, I had this wild idea to revive my personal/professional blog. I have no idea why, but it seemed like a good idea at the time. So I rebuilt the template I had made years ago to go with my website and here we are!
Yes, I do occasionally blog for my job as a Technology Evangelist at Riverbed, but there are lots of other topics I’d like to blog about that don’t seem to be right for that venue, so I’ll put them here. I’m still interested in the Internet of Things, which I used to blog about almost incessantly when I was at Sun Labs (a moment of silence, please), traveling — which I do a lot of these days — and other topics of interest to me. For a list of things I’m interested in, feel free to see this list.
In reality, it could be anything. Being polite company, it won’t be politics or religion, but you can get that anywhere on the internet these days. The last thing the world needs is another place for that drivel. I have no idea how regularly I’ll post, but if there’s something you want to see me cover, feel free to contact me. If it’s urgent, contact me here.
Recent Blog Comments