Made in the USA

Unlike a lot of electronics these days, the green canaries are not simply "designed in the US" - they are made in the United States as well. The software was written in Idaho and the sensors are manufactured in New Hampshire. We try to source parts from the US but the reality is that many components are simply not manufactured in the US at all. I want to give a shout-out here to Adafriut who are a veritable one-stop-shop for most things electronic. Plus they do an excellent job of separating the wheat from the chaff.

Anatomy of a Green Canary

The green canary's brain is based on a Beagle Bone Black board powered by a Ubuntu Linux operating system.

The temperature sensor we use is manufactured by RDing Technology - a Chinese company which specializes in industrial sensor manufacturing.

Connectivity on the sensor side is handled by a Verizon cellular modem. These may vary with time as Verizon upgrades their network and sunsets hardware.

One thing we learned from our field testing is that there is a strong need to harden the "softer" components so they can withstand the reality of day-to-day life in the greenhouse. We enclose the electronics in a hard outer shell. The power cord and the temperature sensor pass through the outer shell. The opening is hydro-insulated with the help of a couple of washers. The opening assembly tightens against the box with a rubber gasket between. The hanging hook is insulated with a watertight washer. As a result the outer case is water resistant (if an errant watering hose hit the box, it will not damage the sensor). Nevertheless, it is worth noting that the green canary is not water-proof. You cannot shove it in a bucket of water and expect it to survive, so please contain your enthusiasm for experiments like this.

In Development

There has been a lot of interest in bringing the sensor out in the field. The architecture of our system allows this (every sensor is autonomous becaues of the individual Verizon connection to the server). The major problem is the power source. While we can power the sensor off a car battery, the cellular modem is a veritable power hog and drains the battery pretty quickly. Nevertheless, we are optimisitic that we can solve this problem (albeit at a cost)

In addition, we plan to add a probe measuring humidity as soon as practical.