What’s So Hard About Building an Internet of Things, Anyway?
February 4, 2015 Leave a comment
- IoT in the media often focuses on end-user devices or rote connectivity such as RFID, ZigBee, WiFi or Bluetooth; only in less-heralded places, like tech blogs, do the “blood and guts” such as mobile connectivity, and the OSS/BSS software and/or professional services needed to stitch it together, get discussed.
- This focus on “things” fails to capture just how much work needs to be done on the Internet.
First things first… shameless pitch time. Last week, I sat down with RCR Wireless for the inaugural episode of its weekly IoT show, Connect This. Take a look; I’ll wait right here.
Good show? Please feel free to share it with your friends.
But, as fun as it was to do the show, and despite the fact that it was about trends in a hot high-tech market, I still found myself feeling like a bit of a fish out of water. The show was about devices and what will drive adoption. All important stuff, but I had to fight the urge to dive headlong into the real blood and guts, not of what will drive IoT adoption, but what will enable it… you know, the Network Matter.
Networking vs. Connectivity
So, we know that, in large part, IoT will be about connected devices rather than networked devices. However, when one wants to talk about a lot of the really cool IoT applications that will take the market to the heights it’s predicted to reach, then we are talking very much about connectivity. Connectivity will drive smart cities. It will drive the connected car, e-health… the really fun gadgets.
For most folks reading this blog, I’m preaching to the choir. However, for those that aren’t knee-deep in networks, networking will be one of, if not, the single most important IoT enablers. Yes, 5G is a way off, even though capabilities such as zero latency and ubiquitous connectivity will be key for most of the really Space Age stuff like all the cars driving themselves. But, what about for now? What about that WiFi hotspot in your Cadillac or BMW, or your pick-up truck. This doesn’t require zero latency, but it does require dense mobile broadband networks. In other words, we need small cells, het nets, carrier aggregation and LTE-broadcast. Otherwise, well, you deal with the Range Rover-driving soccer mom with the three screaming kids because Disney XD hung up while stuck in traffic.
OSS/BSS and Services
Now, if networking might be a tad underappreciated in the IoT lexicon, then the back office is out in left field. Sure, Amdocs, NetCracker – even IBM if pressed – will eagerly talk about the importance of OSS/BSS and services in the context of IoT. And, lest I forget, AT&T launched its connected car program with a strong presence of OSS/BSS players. These notwithstanding, however, I submit that we should be talking even more about the back office. I could drone on about how the sheer number of devices will create a proverbial noodle salad of complexity for the OSS and BSS systems within a service provider, but instead, I’ll let a quote from NetCracker do the honors:
“Each of those devices will have many different attributes that need to be modeled, understood, and inventoried in order to support device activation and service and application provisioning. Further, QoS – both in terms of connectivity and troubleshooting – will have to be managed for every device and all of its connections. Associating devices and applications with customers and users is also a growing challenge. Each of these factors has direct impacts on how enabling OSS systems must be designed and implemented.” – NetCracker
Well said, NetCracker blog. And, let’s not stop there. I want to finish up by talking about not the glue that will hold it all together, but rather the good folks who will apply the glue. Let’s be honest, IoT will be great… as soon as we figure out what people really want, what they’ll really pay for, and then how to actually build it.
This is where professional services come into play. Strategies need to be formed; networks need to be planned, built, optimized and cared for. End-user services need to be supported. Customer experience needs to be measured… and money collected. Luckily, there is no shortage of telecom equipment and IT players to provide these services. Alcatel-Lucent, Ericsson, Nokia, Huawei, Accenture, HP, IBM, Amdocs, NetCracker: I could use up this entire post just naming services players. But, if you’re a telecom network operator, you might need many, if not most, of these names to do IoT right.
Now, what’s so hard about that?