Honestly, I don't know quite yet.
The intent is to design, build, fabricate and deploy various IoT devices.
The first device I'm working on is an Internet of Things sensor for people to monitor their plants.
It's actually part of a broader initiative I've been working on for the last few years. I've been looking at and researching how to cost effectively introduce cheap greenhouses in the Canadian Arctic.
Now, I know what you're thinking: The arctic has no sunlight, nor is it warm!
All true. However, I'd like to utilize technology and proper planning so that the average resident of Canada's most remote territory can grow basic vegetables.
So, the idea started with introducing residents to growing plants at their windows. It's appropriately small scale enough that hopefully, people would be interested in growing their own lettuce, spinach, kale and various spices and other edible plants.
To do this, I need data. Invariably, it's pretty common that people have attempted to grow plants on their window sill using green onions/scallions, lettuce, and various other household vegetables. However, they usually fail. Why? Either not enough sunlight, it's too cold or they weren't being watered enough.
So, I decided to tackle the problem of logging this data to get a better picture of why plants fail. Now, the primary problem is how would you do this? It's easier to deploy technology to log data then it is to hire a scientist or train an intern or volunteer to log this data accurately. How much easier? Well, suppose we did the latter. The data would be inaccurate because it wouldn't necessarilly be continuous. You'd get a brief window every non-determinant period of time, say once or twice a day; so if the plant did fail; you'd likely still not know why because the sampling rate is not near enough in interval to be useful.
I found a potential solution in the form of open source microcontrollers like the Arduino. Unfortunately, they didn't have wifi capabilities. It required wired essentially dumb systems. This is annoying from the perspective of connectivity. No one wants to connect a serial cable to a device to dump data from it because not everyone is a computer/network technician trained in firing up a hyperterminal or putty session and diagnosing the appropriate baud rate needed to dump data. Or to interpret that data by passing it through a software application to make it look pretty. So much work. So little reward. It just doesn't scale.
So, obviously I needed a better solution. This is where the esp8266 microcontroller comes into play. It's an integrated circuit equipped with a wireless chipset built-in. This means it can be programmed to connect to a wifi network of your choosing. Additionally, it's ridiculously inexpensive. You can find these robust microcontrollers on Aliexpress for about $4.
Boom. I found my solution. Now, I have an issue where I need to interface it with something people can actually use. Real people. Not the poindexter with suspenders, a pocket protector and enough time to tinker with a piece of software/hardware until they're blue in the face. The type of person who knows how to use a computer, generally speaking, but doesn't necessarily want to know what assembly language is, what interface is needed to program a microcontroller, how to network it, and other basic administrative tasks most companies and organizations will hire people to do.
So, in order for people to use this hardware, it needs to be simple. How simple? I was thinking you install the app on your phone, install the batteries in your plant sensor(s), press a button and boom, you're already using it because the app takes care of all of the configuration for you.
I'd also like it to be cheap. Why? Because I'm going to be bankrolling a small scale deployment of 50 prototype devices for testing in real world conditions. I can think of no harsher environment than Canada's arctic. I'd say it's a few orders of magnitude less complicated than growing plants on the International Space Station. And before you ask, yes: That's already something NASA is doing.
I'm looking to do the same thing, only on the window sills of homes in Nunavut.
I think it's possible, and I can't wait until it happens.