I read that you decided to only go with the initial size and had put off or possibly canceled larger size Farmbots. Rather than trying to make a Farmbot that is wider or longer or even both, why not instead concentrate on a mechanism that will move the current Farmbot to another planting space. In this way, you could make it take care of 1, 2, or even more plots without a structure change to Farmbot itself but with the addition of something to lift (if needed) and move to the next plot. Of course it would also entail an expansion of the software so it could keep track of what to look for in the different plots. But, as far as planting, weeding, watering and such there is no reason this wouldn’t be a better way than having multiple sizes that take different strength materials to build.
I have seen discussions and plans for a Farmbot that is based of a circle design instead of a square CNC design. http://wiki.farmbot.org/media/farmbots/orbicular_farmbot_1.pdf This would really allow for further square footage expansion without compromising as much structural integrity. The problem that i see with this design is square footage efficiency when downsizing. The current model is great for maximizing smaller gardening plots.
If you wanted to stick to the CNC format, I’d imagine a turntable or transfer table design, similar to what you’d find for locomotives in a railyard, would efficiently transfer the Farmbot from one plot to another. If you had a barcode on each plot, the Farmbot could recognize each plot and differentiate plot from plot. You’d also need the ability for the Farmbot to quick connect to a new water and power source with each new plot. You could mount the vacuum on the Farmbot to minimize the number of fixed supply sources.
Another brilliant suggestion that I’ve seen is mobile planting beds. The Newbs guide to Farmbots limiting factors and strengths The Farmbot stays fixed, but the planting area moves. This could have significant applications in hydroponics!
I can’t wait to see what these expansions look like, once brought to life!
I’m actually working on two new designs, each stating from ground zero. One design will be for green houses and the other will be for growing in a traditional field garden.
Each design will attempt to solve only the “labor intensive” parts of each location. Inside a greenhouse, you don’t have a lot of pests/creatures to get into and eat your crops. You have power, climate and lighting control. From my perspective, the primary challenge is watering and nutrient injection. Nothing else is worth automating. I’m using a modified open source irrigation controller to accomplish this task.
For fields, I think there are really two challenges that automation could logically address. The first is pest/creature defense (think automated scare crow, etc.) and the second is weed control. Again, from my perspective, nothing else is worth automating. In this setting the system must be able to work 24/7 in harsh weather and environmental conditions.
Both the green house and garden field areas require a system that is scalable in size. That’s the final requirement that I see as essential to any successful system that will address this market.
Germinating seeds and planting are such a trivial task and require so little time in both settings that its not worth spending any ergs automating this aspect of farming.
For example, I have a 52’ x 24’ greenhouse and I can “plant” the entire green house in less than 1 hour. In a greenhouse I would never plant seeds directly but would always grow plants in clay pots in a nursery area under a grow light, with heat and proper watering. Then pick the best of the best of the seedlings to plant. Why? Because space is precious and I only want to plant the best plants in that space.
My “classic” field garden is about ¾ of an acre. I have a dedicated well for my garden area with a sprinkler system that can deliver proper water and nutrient to my plants. That is of course automated with well proven, inexpensive hardware. BTW, I’m using the same open source sprinkler controller for the field area as well 
Similar to the greenhouse, I can plant the entire ¾ acre garden in about 4 hours so that’s also a task not worth automating. However, before doing that I have to prepare the soil with my tiller and then hill the dirt with a hiller. This is key in my area as we have a lot of rain during the growing season and without hilling (planting plants on top of hills of dirt) the plants would likely drown when small. Planting gardens on hills is a widely adopted method of gardening in many areas. This step takes about 2-3 hours for my tractor and implements to accomplish and it also requires a fair bit of horse power, something that a robot is likely to not have ( ~50 bhp at the PTO for tilling.) Also note, this implies the ability to get my tractor into the garden area once its been automated.
Once planted, the primary “labor sucking” tasks for a field garden are weeding and keeping unwanted pests/creatures out of the garden, especially as the veggies and fruit start to ripen. Weed control is something that needs to take place year round, not just during the growing season. I plan to use a dedicated and semi-automonous robot (like a robotic lawn mower for example) to automate those tasks. Such a machine is of no value for greenhouses.
Once automated, a 8-16 week growing season would only require less than 4 hours of labor, mostly on the front and back end for planting and harvesting of plants. That means that I will have automated more than 99% of the “growing time” with either an automated irrigation and nutrient system for green houses or automated irrigation plus a weeding/pest control robot for a traditional garden.
Im making a new lawn this year and would like to get a good mower for making stripes. I know I need a mower with rear roller, but id like to do some research into which type is best for me.