Though cannabis legalization is spreading, new taxes drive the price to an average $320 oz. At the turn of 2018, we were interested in developing a smart monitor which integrates with our home automation platforms to make cultivation effortless.
While it's said that cannabis 'grows like a weed,' we've found that environmental factors contribute greatly to the volume and quality.
Not exactly medical grade
This makes home growing high grade cannabis one of the most compelling use cases for precision farming.
Pricey apps offer access to expertise but can't scale. Growboxes offer automated lights/water at the cost of locking the user into a particularly constrained growing environment, not to mention a $2700 price tag.
There really is no environmental control system priced for the small scale home grower, so we decided to make our own.
The modern hacker has an assortment of environmental sensors from the YL-69 for soil moisture, the DHT11 for temperature/humidity, and the TSL-2561 for measuring luminosity.
However, the human gardener primarily relies on visual clues from the plant and growing medium to make decisions on when to water or fertilize.
Computer vision has demonstrated reliability in autonomous vehicles research and we apply this technology in diagnosis/control of cannabis growing environments.
Kindbot is the only garden monitor designed for the modern consumer with home automation platform and smart plug integrations, as well as state-of-the-art techniques from computer vision.
Kindbot is smarter than the competition, empowering the small-scale grower and lowering the barriers to success with:
- control over lights/irrigation/fans via app / chat / Google Home + Alexa VUI
- environmental sensor dashboard
- notifications (text/email/slack)
- yield estimates & pest/deficiency detection with computer vision
- log short notes to your grow journal
In the following, we describe the evolution of the Kindbot home garden automation hub.
Starting with Sensors
Putting a 300W ceramic metal halide lamp into a 3x3 tent will generate some heat so strong ventilation with duct fans and perhaps air conditioning will be required.
The Raspberry Pi Zero measures lumens, temperature, and humidity through GPIO.
With a DHT11 sensor, we were able to monitor temperature/humidity over the wide fluctuations observed from night to day.
A simple process regularly checks the temperature and humidity values and prompts a text or email notification when either falls out of the target range.
These logs helped us to notice that night time temperatures were getting too low so that we could reduce ventilation.
Similarly, introducing a TSL-2561 helped us to optimize the placement of our grow light. Intensity is proportional to distance.
Schematic of early prototype...Read more »