The Internet of Useful Things Still Requires Man to Think for Himself
Does a yellow brick road lead to the IoT?
Frankly, I think the reason why we are hearing the term “Internet of Things” bandied about so much is because the Internet is ubiquitous, and “IoT” sounds more serious than “Internet Stuff.” But the term has been adopted so rapidly and thoroughly that many are tired of hearing it.
In reality, “IoT” is a widely general term. However, in the strictest definition, it refers to devices that have Internet connectivity and operate in a machine-to-machine fashion for communication; that is, no people are directly involved in gathering, or in making local decisions about, IoT-collected data. IoT is ostensibly a bundle of sensors attached to a single board computer (SBC) that uses the Internet to transfer data. IoT is supposed to gather invaluable gems of data from an infinite sea of potential data to do all manner of things, and in the best scenario IoT will improve efficiency, productivity, and numerous lives by predicting the future.
However “garbage in, garbage out” (GIGO) still holds true. This fantastic promise is actually built upon huge amounts of software and hardware development, so we can place a tiny algorithm-based robot wherever some master-thinking-person/team/company thinks it should be located, in order to get the data that said person/team/company thinks would be useful. There’s a Wizard of Oz behind every IoT scheme. IoT is just a computer with a person who programmed it, and therein is the same problem that we have always had with computers (e.g., blue screen of death, buggy apps, frustrating “features,” and potentially lethal security gaps).
IoT may have no “kill switch” for those on the receiving end, and by the time a hack is implemented, it will be too late. Like the guy whose Jeep Cherokee started the AC, tuned to a hip-hop channel, and started squirting wiper fluid, all while driving down the highway at 70 MPH. Hackers were able to remotely take control of his Jeep via the Internet connectionon his infotainment system, to the extent where they finally disabled the brakes and put him in a ditch. This was accomplished via a wireless connection that can be operated from anywhere in the world, and no physical access to the Jeep was necessary, only an Internet connection. Jeep has issued a patch, but the patch must be manually loaded onto a USB stick and installed by sneaker net. An infotainment system should have an air gap to every other computer on the car. Suddenly blasting a radio won’t do more than startle you, but disabling the brakes or the transmission can be deadly, and the driver has no ability to stop the hack in the moment. Scary, yes? Think we can get something for nothing? We do. IoT without security is analogous to getting something for nothing, but at this moment in the history of IoT we are dazzled by the flashing neon promise of what we can do, not what we need to protect.
As for developing useful IoT, some might argue that Artificial Intelligence (AI) will solve the GIGO problem, and that would be clever. With AI, we can sit back and let other computers deploy an ocean of tiny robots to gather the right data to solve problems like World Peace (or Whirled Peas?) while we check out cat videos and text while being driven to the local convenience store for a slushy. (The autonomous, self-driving vehicle whisking us away is also considered IoT by the way.)
Realistically, it’s a chicken-and-the-egg issue that we know the answer to already: at the root of every computer is an army of engineers who makes it happen. Placing a tiny data-gathering computer in the right spot, gathering the right data, to shuttle useful information to a cloud-based decision-making data-sifting machine is art. A good artist knows his media well, and a good IoT engineer will know his even better, because just one successful hacker to IoT is like a vandal who tags Banksy’s graffiti.
“The Cloud” is like The Wizard of Oz, with the great anonymous computing power to make sense of all the data (with the help of a human with a PhD in statistical modelling and several data scientists.) And so there goes the dream of everything magically happening by itself while we pursue other geekly pleasures. And so goes the mystical promise of IoT. It’s like a coffee mug I saw yesterday, a sharp straight pin to my techie balloon that said, “There is no cloud, it’s just someone else’s computer.” A word like “cloud” for a technical abstract construct sounds suspiciously like subtle red marketing herrings to bestow comfort on the unknowing, because storing anything on a cloud requires trust. You don’t really know much about the cloud you are using, who made it, how it’s kept secure, or what happens to the data you think you are deleting. Gathering loads of data without a plan for how to use it is not what IoT is supposed to deliver. IoT is what happens when the self-driving car gathers information about the car’s surroundings (video with image recognition, for example), the most recent map of the area, and local traffic laws (e.g., speed limit) and mashes it up to an immediate decision as to whether to merge faster or slower onto the highway, for instance. Sounds like a computer to me. If they can work out security and the majority of bugs, like the Tesla that, while on AutoPilot smashed into a parked van, I am certain that computers that drive cars will be safer for humanity than humans. Trust, but verify.
Lynnette Reese is Executive Editor, Embedded Systems Engineering, and has been working in various roles as an electrical engineer for over two decades. She is interested in open source software and hardware, the maker movement, and in increasing the number of women working in STEM so she has a greater chance of talking about something other than football at the water cooler.