[davehcker]

Right now (re-starting almost the fourth time), my goal is to get consistent 1. images 2. co2, ec, ph, temperature, humidity, light intensity/quality readings. Anyone putting in money (besides me) hesitated until now because they couldn't see what was in my head. So I'm remaking the setup with data-collection in mind first. It's almost shameful to confess that it is an 8 bed farm only :/ And this will be the setup for next 2 months. Once you've a 1200 plant bed (which is only 10 metre^2 with four layers), I think you will be able to demonstrate the efficiencies with the available data. ps: You can get a harvest every 19 days, roughly.

Computer vision is a big chunk here, and me being no expert, I'm trying to have the hardware setup right so I'd have to do little on the software side of data processing.

If you're doing on your own, and don't want to spend a lot, I'd just say that almost everything can be done with off the shelf boards and sensors.

[nolite]

How mature is this field? I get the impression that all the existing companies doing this have cracked a secret. Can a hacker-type really make significant headway in this space still? (I'm really really interested in this)