Dec 082013

By the end of the decade, self-driving cars will be on the roads in many developed countries. The electric grid will tell our heaters when it is more economical to run, “learning thermostats” will be in many homes, and we will track the movements of people, pets, packages, and many other things. By some estimates, the number of devices connected to this “Internet of Things” (IoT) will pass the number of connected human users by 2016.

The question is: will serious accidents be necessary before people take the risks seriously and harden this infrastructure?

Because the IoT senses and controls physical objects, serious harm can happen — either accidentally or intentionally. We need devices to be authenticated. They in turn need to authenticate the control systems from which they receive commands, and man-in-the-middle attacks need to be prevented – and that’s without exploring the infinite possibilities for industrial spying based on monitoring the signals that traverse this network.

If you make or use Internet-connected devices, you should already be thinking about the identification and authentication of individual devices and control systems, and about the end-to-end encryption of signals and commands. Otherwise, things will go bump (or “boom”) in the night and it won’t be very funny.

[Editor’s Note: This post is part of the annual “Cutter Predicts …” series.]


  2 Responses to “Things That Go Bump in the Night”

  1. Certainly, the infrastructure will need to be hardened, against both accidental failure and deliberate interference. But I think the thing that will drive the adoption of (autonomous) vehicles more than anything else will be insurance companies.

    As soon as the actuaries notice that self-driving cars have dramatically lower accident rates, premiums will drop for those cars substantially (that or human driven car premiums will rise substantially.) That alone will force many consumers to the autonomous car market.

    The other thing will be the transport industries. Heavy transport drivers are the weak point in profits at the moment. They have to rest (by law in Australia) and this slows down delivery times. Self-driving trucks won’t have to, they’ll make the run non-stop, apart from refueling. And who knows, we may even start to see refueling happen while the truck is driving in the same way long haul military aircraft currently refuel.

    But still, you are right, security here will become an issue, but as to your question when or how many serious accidents will it take to ensure the systems are hardened?

    The cynical answer is when the cost of the accidents exceeds the cost of not doing the hardening.

  2. The potential implications of a hacker finding his way into the internet of things are very serious. If we had self driving cars, would it be possible for someone to infiltrate this system and take control?

 Leave a Reply

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>