The Birth of Fae Bot

admin5655DevelopmentLeave a Comment

The Demise of Japanese Rice Fields

When I moved to the Japanese countryside I had a discussion with one of the Hackerfarm founders, Chris Harrington, who also happened to be a rice farmer. He explained to me that, here in the countryside, the rice fields are basically free:  people would give them away if you promised to grow things in them, or at least maintain them, so that that they don't appear to be an abandoned plot as this would be a huge stain on the reputation of the field owner in the area.

This situation is caused by a shortage of farmers who are becoming fewer over the years as the younger generations move to the cities with no interest in farming. Small rice fields also struggle with low profitibility. This got me thinking:  Would there be a way to automate the maintenance of a rice field and turn what is seen as a hassle into a profitable endeavour?

 

Would there be a way to automate the maintenance of a rice field and turn what is seen as a hassle into a profitable endeavour?

Image
Why CNC isn't an Option

We first discussed the possibility of having a gantry over the field, CNC-style. This was a bit before Farm Bot would do exactly that on a small scale. We ruled this out because of the instability of the soil around rice fields and also because of my doubts that installing a gantry of this size could be affordable for a small rice field owner.

Why Wheeled Bots Won't Work

We also ruled out wheeled robots. Rice fields are typically flooded during the year, and their bottom is made of a thin clay-like mud that would engulf any machine stepping on it. You actually need special kind of boots to get into there yourself! Vehicles do exist, but they have very large wheels that guarantee to destroy the plants where they go. They are only used for planting and harvesting.

Image
Image
The Problems with Drones

Drones were all the rage at the time, so why not a drone approach? Well, although there are tons of cool videos with drones doing nifty tricks while appearing very stable and precise, these are usually done indoors without any wind in a place that has been fitted with expensive infrared sensors, that I knew did not like sunlight very much.

Why CNC isn't an Option

We first discussed the possibility of having a gantry over the field, CNC-style. This was a bit before Farm Bot would do exactly that on a small scale. We ruled this out because of the instability of the soil around rice fields and also because of my doubts that installing a gantry of this size could be affordable for a small rice field owner.

Image
Why Wheeled Bots Won't Work

We also ruled out wheeled robots. Rice fields are typically flooded during the year, and their bottom is made of a thin clay-like mud that would engulf any machine stepping on it. You actually need special kind of boots to get into there yourself! Vehicles do exist, but they have very large wheels that guarantee to destroy the plants where they go. They are only used for planting and harvesting.

Image
The Problems with Drones

Drones were all the rage at the time, so why not a drone approach? Well, although there are tons of cool videos with drones doing nifty tricks and being very stable and precises, these are usually done indoors without any wind in a place that has been fitted with expensive infrared sensors, that I knew did not like sunlight very much.

Image

 Still, it seemed an interesting challenge, and I was already planning to try to use computer vision to manage a fleet of small drones indoors.

But then then a new idea occurred to me: Why not try combining drones with a crane for those cases that need high upward forces? As I pursued this idea, I found that the more elaborate the crane design became, the more useless the drones were. I then remembered from my old class in the category of Parallel Robots an oddity named a "suspended robot" or "cable bot." And from this realization, Fae Bot was born.

Leave a Reply

Your email address will not be published. Required fields are marked *