Thinking at a Landscape Scale: Part 1
The building blocks of a planetary sensory system
In the early 2000s, the idea of "fab labs" hit the scene. The promise was tantalizing: micro-scale facilities where you could prototype a wearable fitness tracker, a VR/AR headset, or a new kids toy, and go on to sell millions of them. They held the promise of giving anyone the power to build a new billion-dollar company from their garage.
The rise of the maker movement that also started around that time was driven by the advent of inexpensive microcontrollers such as the Arduino, and ever-declining cost curves have now given us six-dollar, one-square-inch boards that can run machine learning models for computer vision and robotics applications, and single board computers the size of your wallet, like the Raspberry Pi, that can run full-scale manufacturing facilities.
More recently, developments in biological laboratory equipment, processes, and business models have shown promise in the areas of drug discovery and designer genomes. There are growing numbers of bio startups working on big problems with what are essentially desktop, or maybe garage-scale, technologies. Instead of a few massive, NIH- and industry-funded labs, we can have hundreds, maybe thousands of small, entrepreneurial labs working on many more ideas than are viable under the prevailing, monolithic R&D model.
The potential of miniaturization is also reflected in the earth observation industry. Instead of launching one big, billion-dollar satellite loaded with sensors every few years, you send up many limited-purpose, sometimes modular, nano- pico- or femtosatellites.
Finally, the smaller, more precise approach has a parallel in agriculture: precision chemical application, plant-specific nutrient and irrigation delivery, individual weeds zapped by lasers. While these are impressive feats, I want to instead focus on something different: integrating microsensor networks at the landscape-scale.
What do I mean by that? Simply, that nature is slippery. Organisms get out into the wild, and when they get beyond controlled environments, our predictive abilities taper off dramatically. Sure, you can pinpoint a diseased soybean plant with 5-millimeter accuracy, but how are you going to effectively do that on a continental scale?
We still need that big picture, a planetary sensory system, if you will.