After spending a few more weeks thinking about the gardenbot idea, I’ve changed my mind on a few things.
It very quickly became obvious that the idea I had before was too awkward. The scaffolding would block sunlight, and the robots would be too bulky.
After all, when you, yourself, go into the garden, it is not necessary for you to straddle the plants from both sides to take care of them.
So, I started thinking along a different tack, and went back to one of my original ideas.
In this idea, the garden is separated into a grid by walls, and the robots move among the plants by rolling along the walls.
The robots themselves are made of two modules – a base unit which does all the moving, and a top unit which has the specialised tools for managing the various tasks.
The walls themselves are built by a specialised robot.
Essentially, what I’m trying to do is to come up with an “out of the box” solution, where a small group of robots can be set loose on an unprepared area, and build up a functional garden with extremely little interaction from any outside parties.
Obviously, this will take some time to create, but that does not mean that it is impossible.
Anyway – it seems to me that the first robot that needs creation is the base unit – I can build the first walls myself, and I figure that the wall-building robot will also be a top-unit that rolls on a base unit, so the base unit is a logical first.
What I want to do, then, is to come up with a unit within the next year, that can traverse a grid of walls from any point to any other point, carrying something, without falling off the wall, getting lost, or dropping the object.
A nice test would be a glass of wine 🙂 – this would have the added benefit that I could work on developing an algorithm which would describe the optimum speeds for turning and acceleration such that the robot will not drop anything.