Crafting Material Interfaces

Ade Ogunniyi, Fawn Qiu, Sam Jacoby

Nurturing Natural Sensors Kuznetsov, S., Odom, W., Pierce, J., and Paulos, E. Nurturing natural sensors. In Proceedings of the Conference on Ubiquitous Computing (Ubicomp), ACM Press (2011), 227-236.

Ways of Sensing

This paper is not really about sensors at all, at least not what we think of when we think about sensors: electronic that sip temperature, pressure, motion, and light from the environment. “Nurturing Natural Sensors” is about an approach to sensing in general, one which brings a novel perspective on what it means to gather information from an environment. It considers the work of experienced gardeners, beekeepers, aquarists, and others, examining the ways in which they monitor their environments and maintain information about, say, their aquariums and beehives. It discusses the complex relationship that people have with sensing technologies, and the ways in which they use their own ‘natural’ senses, as well as those indicators in nature, to understand what’s going on. This is ‘natural’ in two valences: first, what does it mean to be a natural sensor, and what kind of holistic sensing can we extract from the natural environment?

You are a ‘natural’ sensor

Lets begin with a question: how do we, people, sense our environment? Examine one case: imagine stepping into a crowded cafeteria, say, the stylish one on the 5^th floor of the Media Lab. An alien researcher surprises you with a question, “Quick! What temperature is it in here?” You take a quick look around; most people have their coats off, a few are wearing sweaters, some are drinking coffee. A fellow is speaking loudly, he’s in shirt-sleeves. The air against your own skin feels comfortable. You guess 67 degrees. The alien researcher is astonished. He examines his fancy gizmo. You’re just about on the button. How?

You have compiled a set of instantaneous measurements—the amount of clothing people are wearing, the liquid state of water, the absence of vapor puffing from open mouths—and given an accurate estimate based on your experience. It’s easy. We all do it every day.

This is how we sense our environments and for a digital sensor, it’s a very difficult thing to do. We make use of our perception in a broad, holistic way, stitching together observation, experience, and our own ‘five’ senses, to come to rapid and (usually) accurate conclusions. This is very different than the way in which digital sensors work.

You could, of course, have turned to the wall-mounted thermostat and gotten an exact temperature reading (68 degrees), but you wouldn’t do that—and you don’t have to. We typically rely on our experience over technologies—unless, perhaps, you notice something deeply unusual. Perhaps everyone is wearing a ski-jacket and coffee spitting from the machine tinkled into ice as it hit the bottom of the mug—something absolutely counter to all of your cafeteria-related experience.

What temperature is it inside? Freezing!

Temperature, of course, is relatively trivial. Measuring it is something that we all have a lot of experience doing. Not so, with, say, the health of an aquatic environment, or the behavior of bees.

In ‘Nurturing Natural Sensors,’ researchers spoke with farmers, zookeepers, beekeepers, and others, who make complex judgments about natural systems. The participants gauged the health of an ecosystem by integrating information from many sources; their own observations, their long-time experience, and of course, their senses. They work like us, gauging the temperature of the cafeteria, filtering many indicators through their experience, and arriving at a conclusion. These indicators, or “biomarkers,” come from the natural world itself.

Biomarkers

Telling the temperature in the room by looking at people’s clothing is easy. We’re all experienced at doing exactly that. What about in other contexts? How would you gauge the conditions in an aquarium? The health of a beehive? What does a cold fish look like? All natural systems have biomarkers that relay information about their condition as long as you know where to look. This is where experience, education, and close observation are important.

In ‘Nurturing Natural Sensors’, the aquarist observes the color of the water in a tank, and deduces the corresponding environmental conditions: “I can look at this water and tell that the ozone system hasn’t been working for two weeks on this…because it’s green.”

Sweet Fishtank

This is a one-to-one biomarker. Green water in the fish tank means that there is an ozone deficiency. Such a condition, of course, could also be confirmed using a conventional sensor—and perhaps might be, later, to get a better sense of the particular situation. The canary in a coal mine, of course, is an example of a one-to-one biomarker. Miners, working in hazardous conditions, would use a canary’s behavior to gauge the relative safety. The birds, being more sensitive, would become sick before methane and carbon monoxide affected the miners themselves. This is a classic example of an ‘animal sentinel’, in which a more sensitive organism’s reactions are leveraged to expand our own.

Canary in a Coal Mine

Other biomarkers are more ambiguous but no less informative. Fish that are ‘sliming’ heavily are a sign that something is amiss, “it could be water quality…a parasite.” This is a one-to-many biomarker. One indicator could have many different causes. Integrating many such indicators, observers can draw broad conclusions about an entire ecosystem: a beekeeper observes the amount and color of nectar that bees return with, drawing conclusions about drought or “when the first flowers are blooming in the spring.”

Of course, participants in the study do not depend on their observations alone, and digital sensors augment their observations and guide them, particularly in moments of uncertainty or when more accurate information is needed. Their relationship with those tools, though, is not simple. The researchers observed that “over time, participants developed a ‘zen feel’ for the complex processes in their environment.” Participants were able “readers” of their respective ecosystems, synthesizing broad conclusions about its overall health from subtle cues. This process points to a different way of thinking about sensor systems, and the ways in which people use them.

Natural Sensor Systems

In “Nurturing Natural Sensors,” the researchers propose a different way of thinking about sensor systems. The technologists’ approach, they argue, is too limited in scope, excluding “living organisms and traditional tools” that could have a valuable place in a sensory framework. A great deal of sensing technology is focused on automating routine tasks and making human oversight unnecessary. No one would argue that this is not useful, but the researchers argue that this kind of focus overlooks a great deal of opportunity in the research space: sensor networks could be constructed to work much more like people do, “integrating inputs from plants, insects, and animals along with analog tools and digital sensors into holistic representations of the environment.” The technological problems of incorporating a canary (as opposed to a chemical sensor) into a sensing system aside, this opens the door to a new set of considerations:

• How do sensors affect the ways people interact and experience environments?
• What kind of learning and understanding do these interactions promote?
• How can natural organisms ‘output’ be integrated into a sensor network?
• What is the human role in a responsive sensor network? Is there one?
• Does ‘smart’ technology make us smarter?
• How do we encourage “human awareness, reflection, and wonderment about our living world.”

Other Research

“Nurturing Natural Sensors” was published just over a month ago, in September at Ubicomp 2011—and it explores a relatively theoretical space. The most interesting and relevant reading is Paul Dourish’s Implications for Design. Other interesting, if tangentially reading is W. Willet et al.., Common Sense Community: Scaffolding Mobile Sensing and Analysis for Novice Users, and Y. Rogers et al., Ambient Wood: Designing New Forms of Digital Augmentation for Learning Outdoors.