What is the Internet of Things?
Simply put, it is when everyday ‘dumb’ things like your shoes and swimsuit start sending (and possibly receiving) data to or from other devices, wirelessly I am sure, giving information relevant to their status and function. When this communication ties into the Internet, these things can form a web of live data sources known as the Internet of Things. But let us not get ahead of ourselves and get some background first.
We have come to understand what “smart’’ means in smartphones. it’s about being loaded with features typically found on laptops or bigger machines, and even features that once required its own hardware like Navigators and high resolution video cameras. In 2013, mobile phones accounted for 17% of all internet traffic worldwide according to StatCounter.com. It is expected to increase substantially as more heavily populated developing countries improve their Internet infrastructures and reduce connectivity cost. But what does that have to do with The Internet of Things?
Mobile phones have driven leaps in multiple technologies across hardware and software; from miniaturization, cheaper memory, lower power consumptions, to efficient algorithms that offload much of the processing from backend servers to the phones themselves. This mobile device that most of us are attached to, has become a very valuable feed to the rest of the world about our micro environment; reading, measuring, and, when we choose, sharing this micro environment with our private network of friends. This in turn spurred innovations in all types of micro sensors that smartphones makers can add on, sometime as external accessories, to their devices. Sensors like motion measurement, human voice detection, optical sensors, temperature, pressure, acceleration, body vital signs, and of course GPS and altitude to name a few. The advances also added intelligent and efficient software like speech and facial recognition, noise filtering, even machine learning algorithms that automatically build models of what to expect from a sensor and take action when a new reading does not fit the expected input. All of these new innovations were made accessible to everyday consumers at a very reasonable price that is getting cheaper by the day. The lower cost was mainly driven by fierce competition to tap into the smart devices market. It is also because of the mass production of what is known as System-on-a-Chip (SoC) hardware components. These components integrate complex functions in a dedicated microchip and perform them much faster than if done by the main computer processor. These SoC’s created a Chinese buffet of capabilities for device designers to assemble from. In the last couple of years, consumer electronics and appliances manufacturers have been tripping over each other to jazz up their devices with these components to get their customer’s attention and hopefully add incremental value over the competition.
We are now approaching a critical mass of low cost technologies coming together to remove the human from the center of interaction. In other words, just like a car manufacturer can remotely monitor your new car and call you to come for repairs before the engine light flashes, your fridge can text you a list of its contents just as you are entering the supermarket door so you can decide what groceries to buy. Or a shopping cart that adds up the prices of the items in it before you stand in the checkout line. Or better yet, skip the line altogether and wave your credit card as you exit the store and your receipt instantly shows up on your smartphone and logged into your expense ledger. These are not futuristic fantasies, these are current and proven technologies which are not widely implemented yet.
This availability of sensory data from “things” has created a new technology paradigm. Every day functional consumer dumb things like shoes, door keys, swimsuits, pocket wallets, handbags, pens, and school notebooks are candidates for these sensory microchips that “call home” via the nearest Wi-Fi and summit reports to a master program with built-in intelligence to make use of it. This master program is likely running in a secure cloud that has a big mouth to swallow gigabytes of data per second. The idea is, by having all this data in one place, the human who owns these “things” can delegate some management tasks to a software agent to detect events, trigger responses, or just provide alerts and recommendations on what action to take.
The Big, the Messy, and the Meaningless Data
Enter Big Data. You can imagine now if a manufacturer wants to use this technology for its product lines, there has to be an existing infrastructure to handle it. Possibly a “Service Provider for Things” to which the products can connect once they are registered or activated. The owner of these products can then monitor the generated data and possibly select a software to process it, or maybe turn it off when she pleases. Take for example a Nike walking shoes equipped with this technology. It can read a great deal of data every second: your pulse, forces of impact, pressure points, the relative positions of your feet, your location, foot orientation, speed, etc. Now assume Nike will sell a million of these shoes worldwide and you immediately start to see the volume and velocity of data coming out of them. You would think that Nike or some other player must have prepared the grounds to deal with this outflow if there is a value to be gained from it by the end user or Nike itself. We are seeing some of these preparations in infrastructure building up in cloud computing but we have a long way to go yet.
If you are one who likes to surround herself with smart “things” like the Nike shoes above, then your “things” are creating serious network traffic. Collectively, your “things” are now generating massive data that may overlap, send readings that contradict each other, report all sorts of bugs or errors. This becomes a lot messier when the software running at the “Service Provider of Things” attempts to integrate data from your smart things for your personal “Dashboard of things”. Your shoes can be reporting that you are enjoying a walk at 4 MPH at the same time your swimsuit is experiencing a 150 F temperature water. If the software was design to assess your daily exercise activities, it has to decide: Are you waking or are you sitting in a hot tub? Could it be that you put your swimsuit in the washer and went for a walk? Or is this just meaningless data to skip over? Maybe it is time for the master program to interact with you to learn how to deal with these inconstancies. It could text you a question: Hey John, I am confused! Are you walking or in a Hot tub?
A new field of Big Data science has just opened up as you can see. A combination of many disciplines in computer science covering communication protocols, massive data management, machine learning, and user interface to name a few.
The Road to Reality for the IofT
So the technologies are obviously there, cheap enough, and tiny enough to deploy. Where do we go from here? Like an army of well-trained soldiers, equipped and ready to go to battle, these technologies need a great general to lead them. Someone with vision, but also loved and respected by the field captains and the populous alike. Someone with a vast network of partners and supporters that can lend skills and expertise at a twitch of a finger. Someone with massive wealth to spend unabated until the war is won. Of course no one can be that good. However, someone has just stepped into the ring that may have a real shot, Finally!
On June 2nd, 2014 at Apple Worldwide Developer Conference in San Francisco, Apple signaled a clear intention to enter this realm by announcing two programming kits for developers to build on: HomeKit and HealthKit. These two kits will work with Apple’s upcoming release of iOS 8 and are meant to help developers build Apps specifically for “things” at home and for your health and fitness.
The HomeKit allows programmers to write Apps integrating Apple devices with other smart home devices like Smart TVs, Furnaces and Air Conditioners, Smart Refrigerators, Electricity Panels, Security Systems, light switches, and other devices that may come along in the near future. This announcement is significant not just because it comes from the dominant leader in the smart devices market, but also because it is backed by a real functioning infrastructure that promises seamless integration into an existing and well-functioning, although proprietary, “Internet of Apple Things”. The kits are composed of code libraries and instruction manuals to developers describing messaging protocols and how to command and receive responses from other smart devices. This opens the door for mass participation from developers, but more importantly it creates new expectations from end users forcing developers to innovate and investors to back them up.
The Internet of Things is not real yet, at least not the way my buddies and I used to dream about it when I was a fresh programmer 25 years ago. I told my programmer friends then it won’t be long before we are wearing smart baseball hats that read our thoughts and let us communicate our ideas in perfect articulation to whomever we wish without moving a muscle. A hat that quietly reminds us where we left our car keys and the time of the meeting we all try to forget but have to go to. Well, I was wrong about the smart baseball hat, but smartphones and the eyeglasses from Google are a step closer to my dream.