Wireless sensor networking technology has moved from hobby circles to interconnected consumer electronics, appliances, and devices for the home. In the following interview, “Building Wireless Sensor Networks” author Robert Faludi (@faludi) discusses the practical application of sensor networks and how he thinks they will evolve to meet a variety of needs.
For the uninitiated, what’s the appeal of building wireless sensor networks?
Robert Faludi: Wireless sensor networks are distributed, therefore they’re good at seeing things you can’t. For example, if you place a single soil moisture sensor in a field, you’ll know the moisture at that one point. But if you place 100 of them you’ll be able to learn the entire topography of the moisture in the soil, whether it’s the same everywhere or different, and whether those differences change in shape over time.
If you’re interested in urban pollution you can measure at ground level, at walking height, and at the height where a child breathes. This reveals the distribution of danger across an entire landscape — and you can see that across a week and across the seasons.
Because the networks are two-way, we have the ability to distribute actuator networks of devices that can physically affect the world based upon the special things that sensor networks can see. If you think of a laptop computer as a processor of local data, imagine these sensor/actuator networks as massive distributed devices, kind of like those <a href="enormous underground mushrooms in the Pacific Northwest.
To me though, the scale and economics are secondary to the greater purpose of conveying the stories that this distributed data has to tell us, and improving our lives by employing those lessons in the same distributed way.
Has interest in wireless sensor networking reached a tipping point?
Robert Faludi: I think we’ve crossed the chasm from academic experimentation toward widespread commercial deployment of low-power, low-bandwidth wireless sensor and device networking. Smart energy, home networking and industrial implementations are well out of the gate. At the same time there’s an astounding opportunity for proving the value of these networks in the market.
Imagine the Internet around 1994, with a few high-profile deployments and limitless headroom for individuals to create something and get it online. Device networking is in a similar place. Those with a knack for the technical and a sense of adventure can create the killer applications that will drive an inventive industry for years to come. It’s getting to be a pretty exciting time.
What technical skills or background do you need to build wireless sensor networks?
Robert Faludi: People who have used a microcontroller platform like Arduino, or who have tried their hand with an accessible development platform like Processing, will certainly have a faster path. Really, if you trust yourself, follow instructions moderately well, and can be patient while you learn, that’s all you need. I’ve taught middle school students the basics and they built the book’s first full project in a single 45-minute class. So it’s not so much about your skill level as your willingness to try something new.
How do current sensor networks and technologies need to improve?
Robert Faludi: These components are already quite cheap, but it would be terrific if they were even cheaper. As they become more widely used, a radio that costs $18 will eventually come down to $1 or even $0.50. Low cost is important if you want to fool around with a network of 100 devices.
Range, stability, and setup will also improve. The dream is to just turn the radio on and have it figure out which network to join, what its role is, and configure itself. That’s possible right now, but you have to do a lot of advanced planning. I’m hoping that new technologies will keep making this process easier, so that in the future nobody needs a book like mine to get the job done.
What lies ahead for wireless sensor networking?
Robert Faludi: I’m looking forward to a future where wireless communications are part of everything around us. It’s not about the cool factor — though it would certainly be cool — but more about creating behaviors and interactions that can be carried out in an intelligent fashion. I want systems that help free me up from menial tasks like setting a clock or punching instructions into a microwave or endlessly silencing falsely-triggered smoke alarms. Devices are smart right now but they can’t talk to each other properly, so all the things they know and detect are kept hidden from other systems. The great systems that people design are going to be the first steps in making that cool and useful world a reality.
This interview was edited and condensed.