Plantoids Pilot: Building robots at the Botanical Garden

About time for a report! *salutes*

After some time in stasis, I was able to pick back up on the Practical Plantoids platform eariler this year, in late winter. I dusted off the prototypes, saved my pennies for a refresh of updated gear for the Plantoids. I made a a starter version of the plantoids that any beginner robotics enthusiast could assemble and wrote a suite of Arduino sketches to experiment in several ways with the concepts we're driving toward with a Plantoid. These don't carry a direct connection to MRL *yet*, but when you up the computing part of the brain i.e ODROID and Raspi 3 hardware, you can directly control all of these Arduino functions easily. I'm hoping to get there and finish a more research oriented robot, but I'm also happily making a batch at a time and working with my local botanical garden to have workshops in which we build lots of basic versions of plantoids and exploring the concept. I'm going to launch a batch of kits on kickstarter very soon, and hopefully a success there means I'll get a little more time and resources to play with bigger, more complex versions of the Plantoid. 

With the increase in volume, keeping a nursery of well chosen specimen is fun and challenging. I see the need to make a robot specifically to tend to the carnivorous butterwort babies while they wait to be planted as brains to the next wave of diminutive robot explorers. 

Its has been fun to realize that these living plant specimen are like the great explorers of their kind. They rise from humble plants to the first of many to be given to science and experimentation, all with the increase of their prosperity and experience in mind.



Code has been on my mind for some time, and I'm doing my best to optimize the functions at hand in Arduino to use every sensor to aid in behaviors that help keep our specimen rolling and spelunking safely. In workshops, I've discussed this as the "orders of awareness" aka survival priorities for the plants. From top priorities like immediate obstacles, the rangefinder always must sample its sensor first. Other data points such as light and water are not immediate reaction requirements, but are a matter of life and death, with water then light being the next priorities in line. Each sensor must have a variable stored that will help trigger state changes and reactions. Sometimes the current code gets stuck in a loop and bumps into things too much, but hopefully I can iron out those issues here. 


So, I expect to have this simple Plantoid Robot released in a batch of kits very soon, please stay tuned for that. 

Attached is the code that I've written for the plantoids, which is still undergoing testing there. If anyone would like to help edit and optimize the code, please do so!

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
GroG's picture

Great to see the Plantoid

Great to see the Plantoid back. I was scared they were yet another extinct species, but now I rejoice in the new wave of plantoids being set free in the wild.

Have you any esp8266 ? It might be a nice brain selection. Small, inexpensive, wifi & i2c capable. The wifi capable has made a few people eye the connectivity to a MRL instance on a different machine. Such that you could fire up a central MRL instance and all your drone plantoids with esp8266 brains would dial in - give reports "HAIL BEPSL!", and take commands. No one currently has created a solid bi-directional pipe, but Mats & Calamity both have experimented in this. And I know its possible.

Excited to see this generation !

DJUltis's picture

I was gonna...

Slap an RPI Zero W on one and solve most of the problems there. Then WebGUI can serve as a control UI I hopes. 

A little bit bigger version is on the way. I'm really wanting to build a nice set of robot legs again.