Whenever the Internet of Things (IoT) gets mentioned in the press, there is a sense of an awesome monster about to be unleashed on a poor, unsuspecting IT industry. Billions of devices, more big data than we have ever seen (maybe we need to call it bigger data), and computing up the wazoo!
Is it really that way? An unbounded cosmos of newly connected things monitoring our health, checking our every move, and knowing how many calories we had for dinner three weeks ago? Indeed, the idea of unlimited data, unlicensed and perhaps out of control, is scary for anyone in IT.
However, portraying IoT as a vast river of information is a bit simplistic, to say the least. Most of those connected devices are going to be bought and paid for, so one assumes they have a value to someone. The owner will want to control the data the things generate, milking it for value. This isn’t a model of rivers; it’s a model of many streams.
As an example, a retail store may have dozens of display monitors dotted with blue-light specials. These are going to be driven by a big data application that figures out who is nearby, from cellphones or facial recognition, so they’ll be talking to the company’s computers at a datacenter. Every store will have connections, but there’s no way another retailer will get access to the data streams.
That model of separate streams will repeat everywhere, and the implications to IT departments are that IoT big data, for them, will be bounded and reasonably sized.
That’s not to say it won’t be large. The end product of many IoT streams is a gem of knowledge, and knowledge is fungible. Those buying habits may get sold on to other retailers or marketing companies, or to vendors of goods. These add to the big data flow, perhaps by considerable amounts.
How is all that data going to move around? The IoT is going to be almost exclusively wireless for the simple reason that "things" will cost mostly in the low hundreds, while wired connections cost over $500. The flexibility of the wireless connections from a positioning point of view also counts for a lot.
This means a lot more WiFi bandwidth will be needed, which will need to be backed by good WAN links. Shopping malls might want to think of fiber connections, for instance. This is going to ruffle a few feathers at the big telcos, at least in the US, which publicly state that DSL is fast enough for everyone.
That’s easy, right? We know how to encrypt? The truth is IoT is treacherous ground in this area. As anyone at Bletchley Park would have told you, repetitive short messages with the plaintext in a fixed known format are easy to break, and that’s what IoT will be if we are not careful. Here's one potential scenario: A door lock says, "I’m shut" every minute from midnight to 7:00 a.m. It would be easy to figure out the encryption and to send a message to unlock the door, or of course, you could just sense the guard opening the door in the morning.
It isn’t just door locks that need consideration -- medical implants will be tied to the IoT. From RFIDs on artificial hips and breast implants to drug dispensers and heart monitors, wearable tech and implants are vulnerable to the same hacking problems, but the resulting mayhem could be very serious.
Still, IoT isn’t a frightening giant ogre. If we stop admiring how big it is and realize the devil is in the details, we should be able to handle IoT just fine. The IoT approach will make deploying smart gear much easier and can improve our lives, but we must knuckle down to on issues such as security and wireless bandwidth before we have a billion IoT defibrillators deployed.Jim O'Reilly was Vice President of Engineering at Germane Systems, where he created ruggedized servers and storage for the US submarine fleet. He has also held senior management positions at SGI/Rackable and Verari; was CEO at startups Scalant and CDS; headed operations at PC ... View Full Bio