Over the next decade or two we can expect to see general purpose computing, sensors, and wireless networking, all bundled up in millimeter-scale sensor motes that can drift in the air currents around us.
As processors get faster and smaller we seem to reach points of inflection, our technological progress allows us to do something that wasn't really possible before, or at least wasn't practical. We've seen this before: computers the size of rooms gave way to ones that fit under the desk. In time those too gave way to slabs of aluminium and glass that we carried around in bags, and then to computers that fit in our pockets, and just happen to also be able to make phone calls.
Known as Bell's law, this proposition is intimately intertwined with the far more well known Moore's law. But while Moore's law may be dying, Bell's law, at least for one last turn of the wheel, seems to be holding. You might think the change I'm talking about is the arrival of the much hyped Internet of Things, but the connected devices we see today are really only the first clumsy steps towards something else. They're transition technology.
Over the next decade or two we can, perhaps, expect to see general purpose computing, sensors, and wireless networking, bundled up in millimeter-scale sensor motes that can drift in the air currents around us. The dust around us will soon become smart.
If you think this is just a technological fantasy, it's not. As William Gibson said, "The future is already here — it's just not very evenly distributed." The Michigan Micro Mote is the world's smallest computer, measuring just a couple of millimeters it currently comes in three types able to measure temperature, pressure, or to take images.
Connected devices are a transition technology.
Of course, the idea isn't new. Kris Pister at UC Berkeley coined the term "smart dust" in the mid-nineties as part of a DARPA research project, concluding that, "We will program the walls and the furniture, and some day even the insects and the dust."
If we think ahead to the time when everything is smart, and everything is networked, when computing has diffused out into our environment, the phrase data exhaust will be not longer a figure of speech but a literal statement. Your data will exist in a halo of devices surrounding you, tasked to provide you with sensor and computing support as you walk along. Calculating constantly, consulting with each other, predicting, anticipating your needs. You'll be surrounded by a web of distributed sensors, computing, and data.
We're on the edge of "no sparrow shall fall" computing, and inevitably the architectures and protocols of today's Internet of Things connected devices will heavily influence the emerging smart dust architectures in just the same way the protocols and architectures of the other Internet, the digital one, has influenced the the Internet of Things.
The real question then is who will have access to the sensors, the computing power, and to the data that they generate. Whether the architectures for the smart dust networks will be peer-to-peer and make that computing power and sensing available to individuals, or whether the network architectures will centralize command-and-control into a few hands.
The diffusion of computing into the environment will mean not just that computing power is always available, but that this computing power offers the possibility of continuous monitoring and surveillance. If you think about the kinds of information that today we approximate, determine by polling and sampling, and then by making a calculation. In the future, we won't have to calculate; we'll measure.
The prospect of smart dust brings the arguments around the Internet of Things into stark perspective. It makes the lack of concern around the security of our things terrifying rather than just worrying.
What most people are missing about the Internet of Things is that sure, we can connect a bunch of things to the Internet. But it's not about that. The prospect of smart dust makes it clear, the Internet of Things isn't about connecting people to things, or things to the Internet; it's about putting computing and sensing everywhere. We need to build and design our things for that, because that's what's approaching. Faster than you might think.