For two and a half years from when I had the initial idea, FarmBot went by the name of TRAG systems, or Tracked Agriculture. During this time, I completed many hours of leg work including research, brainstorming, sketching, modeling, prototyping, and reaching out.
This work is the foundation for what I envision FarmBot to one day become and ideas are still helping to guide development today. Below are some of most important milestones and revelations I had during this time of development and each bullet is linked to a post going into more detail.
There exist two types of efficiencies in food production: biological and machine. The efficiencies have classically been more or less exclusive, with biological efficiency coming with a polycropping paradigm and machine efficiency coming with a monocropping paradigm. TRAG/FarmBot is the first technology and method to get the efficiencies of both paradigms.
- Precise, repeatable positioning of the tooling in relation to the ground and plants is perhaps the most essential and defining feature of the system. This opens up the possibility for many unique operations such as precision watering, nutrient mapping, and selective weeding.
- Traditional cubic packing structure of plants is no longer a constraint due to the following three reasons: the ability to place plants anywhere in the x-y plane, the fact that the tooling implements come from above, and the ability to grow different sized plants at the same time. The change to a hexagonal packing structure allows for a 12% increase in space efficiency to 90.7% and non-standard packing structures with different sized plants could increase this number even more.
- Designing for scalability is a major driving force in hardware and software development. The goal is to use the same software for any size system and to be able to continually expand and scale the hardware.
- Plant profile data (how to grow each type of plant) will be public data, easily downloaded into the system for quick setup. It will also be changeable and shareable on some sort of wiki/data repository.