Student project "AWR² (Autonomous Weed Removal Rover)" : we need your opinion !
ld160183 wrote 01/09/2020 at 14:22 • 1 pointHi, makers!
As part of student project "AWR²" (Autonomous Weed Removal Rover), we would like to design a "robot vegetable garden" capable of maintaining an urban garden; this project will then be made available in open source. This robot will be able to move autonomously in the vegetable garden, to map its territory (finding the best path through vegetation and uneven terrain) and to detect weeds (thanks to its camera allowing image analysis by neural network and for the purpose of a weed control option).
We would therefore need your opinion! If you ever own a vegetable garden, do you think such a robot would be useful/relevant in your garden? If you don't have a vegetable garden, would this project encourage you to have one? And finally, since this project will be available in open source, would you be willing to put 2000 euros (component price) to build this robot yourself?
Thank you in advance for your answers!
Signed all the team of AWR²
Discussions
Become a Hackaday.io Member
Create an account to leave a comment. Already have an account? Log In.
hi guys. From my research on farmbot, it seems that the strategy against weeds is: 1) have a map with your every plant. 2) kill everything not on the map. I think there is a big room for improvement there, ie, CV-based detection.
I have been pondering an experimental setup that does away with a rigid farmbot-style frame in favor of steel ropes and strings. What i'd hope is that this hardware would cover a big area very cheaply, and the software would be made smart enough to work with the inaccuracies.
And those 2K EUR + assembly + setup? If that's the price for which i can set this robot free on a 1500m3 area and only check on it on the weekends, it's worth it. But then the risk of theft would be a bit of a concern around here, so again cheaper hw, smarter sw maybe makes more sense?
Are you sure? yes | no
[this comment has been deleted]
Thank you very much for your support ; we hope that our robot will be a success. We will keep you updated !
Are you sure? yes | no
You might want to check out https://farm.bot/ . They have been developing and selling an automated system for some time and have about 1000 installations. There system does much more than weed.
Are you sure? yes | no
Thanks for your advice ! We know and admire the Farmbot robot : we've already tried to get in touch with them. With our project, we wanted to make a more scalable (to every gardens) and accessible robot.
Are you sure? yes | no
Agreed $2k seems too much for a typical vegetable garden; also in many gardens there's the challenge of navigating multiple raised beds. But a roamer that could crawl the whole yard pulling the most common weeds would be awesome.
Are you sure? yes | no
Thank you for your response ! The goal of our robot is that it can move in all types of terrain (why not in a courtyard!). The price is for the moment only a ceiling estimate, and we hope that it will be feasible with less means.
Are you sure? yes | no
I would look at the weedinator project on hackaday.io. They are looking at a slightly different market to you but there is a lot you could learn by comparing and contrasting their system choices with those that your team thinks would best suit your application.
You might consider a market pivot and look at eg a microgreens or niche ingredient farm for the food and hospitality sector. They may be more willing to invest in a robot to increase productivity. And they will probably be able to give requirements that are easier to distill into a design brief.
Are you sure? yes | no
Thanks for your precious advices and your response ! We didn't know the weedinator project's existence but we will look at it.
Indeed, being part of a specific market could allow us to differentiate ourselves. The goal of the project is that it can be adapted with any database afterwards so why not microgreens!
Are you sure? yes | no
The vegetable gardens that I've had myself have been too small to warrant an outlay of €2000 (US$2200) for this sort of thing. However, the house that I grew up in had just enough garden space that, if I lived there now, I might be willing to do. If the robot could also roam the lawn to find and pull weeds, then that definitely would be worth the price (especially given my own physical limitations/disabilities).
Are you sure? yes | no
Thank you for taking the time to answer us!
The 2000 euros is a ceiling estimate and we hope to be able to do with less. Indeed, the robot will be adaptable to all types of gardens (including in the lawn to find and remove weeds!).
In any case, we appreciate your interest and your opinion!
Are you sure? yes | no