| December 24, 2023
by the Open make team and Rory Aronson. Copyright to the authors, distributed under a CC-BY 4.0 licence.
Banner image: farmbot logo, distributed under a CC-BY-NC.
Interviewee: Rory Aronson
Interviewers: Robert Mies (TU Berlin) & Moritz Maxeiner (FU Berlin)
Transcription and editing: Diana Paola Americano Guerrero, Robert Mies, Fabio Reeh, Moritz Maxeiner & Julien Colomb
Screenshot of the interview.
Photos of Farmbot in action, by FarmBot Inc, CC-BY
- Main website: https://farm.bot/
- 2023 White paper : https://farm.bot/pages/whitepaper
- Project start: 2013
- Core development team size: 3-5 paid people + around 5 external contributors.
The FarmBot is basically a large, outdoor resistant, 3D printer for the garden. Instead of extruding plastic, it plants seeds and sprays water.
It is a full manufactured product.
The reality is that most people want to buy the machine from me.
Two companies tried to reuse the spec, but failed mostly because the manufacturing quality was not sufficient.
I essentially wrote a white paper like a senior or capstone project. It’s about 50 pages long, it outlines at a very high level the concept how the hardware could look like and roughly what it would cost.
FarmBot has become a default hardware choice for a lot of those researchers. They need to do some very specific specialized stuff for their experiments and FarmBot is very much useful because it’s open source and it’s well documented.
A lot of that knowledge for farming and gardening is kind of inaccessible unless it’s your profession or major hobby. There’s not a lot of open source, community created or community shared agricultural related projects out there.
How did it all start with the project FarmBot?
I thought where’s the CNC machine or 3D printer for the garden? […] I googled around and didn’t find such a machine. I started working on the concepts after I graduated from school. That was in 2013.
I researched how everyone else in the agriculture industry does things.
I published this paper, it’s all open source. I asked who wants to help me building this.
You’re referring to this kind of intergenerational knowledge, this word that comes traditionally with farming, isn’t it?
From a technology point of view, 3D printer indeed are similar to CNC, the analogy is there.
In the early days it was very much assessing all the people who were getting in contact with me. I empowered them with resources and told them what we’ve got so far. “If your skill set is with Arduinos and writing microcontroller software, here is some research, documentation and a GitHub repository. Let’s hop on a call and and figure out some features, framework or something that you can help build. We’ll just go from there.”
I was orchestrating all the people that were interested and gathering their contributions into one central space and empowering somebody else who came along with the information they needed to make a meaningful contribution.
Could you describe the overall process from how you evolved out of a student project to making a company? How is that structured around the idea of community?
It is really important to go from not just open source but useful source.
If my business fails tomorrow, the vast majority of the work we’ve put into this business is available and preserved. It’s still very valuable.
Could you classify the different products in terms of mechanical, electrical and software?
The FarmBot is basically a large 3D printer. Instead of extruding plastic, it sprays water.
On the software side, there’re three primary code bases. (web application, FarmBot OS which is an operating system that runs on the Raspberry Pi, Arduino code base).
Being a useful source means reducing barriers to use that source content for something useful.
Do you need previous knowledge to understand the documentation?
Do you need to pay for Onshape if you use it commercially?
When I published the paper online, it was revolutionary in a way.
A major component of the paper was the idea to make Farmbot, the concept and eventually the company, all open source.
Our work in progress is only shared amongst people who are actually working on that next feature or version of the FarmBot. That’s our internal team and maybe some small number of community members who are helping us doing some things.
We’ll do the big publish and public release when the product is publicly available.
A lot of being useful is about choosing the right tools for internal development and external sharing with people.
If somebody emails us with a problem, our support policy is to update the documentation and send them the link.
They can come to the forum, search for certain problems and find the solutions on the forum. It’s about having the right information in the right places and formats.
You want to direct people to the information that they actually are looking for.
What information do you share in the end?
The software information is on GitHub mainly or where can I find it?
How important is the scientific community for your business? Is the scientific community part of what you’re doing today?
How many FarmBots have you sold over the years?
They get the base machine from us and they make modifications.
There have been two other FarmBot businesses where somebody wanted to sell the hardware. Both of those businesses failed (It was cheaper but there were problems and the quality wasn’t as good).
Do others have independently started their own business with this information you provide? Have others made their own FarmBot without calling it FarmBot?
Could you benefit from more skills to the ecosystem because of other people creating a similar businesses?
There was the Shuttleworth Foundation. They provided a grant in the beginning over $125,000. Then, we had an initial crowdfunding campaign, It’s a preorder crowdfunding model. The reality is that most people want to buy the machine from me.
How did you start with the FarmBot? How did you get some funding after you were a student?
We don’t publish anything until it is ready for the public.
We try to not share that because it will confuse people more than help them. We’ll do the big publish and public release when the product is publicly available.
[The source is] normal product information with assembly instructions and troubleshooting stuff along with all the information you need to make modifications to the machine or to build it yourself. With one engineer and one software engineer a team can build a working machine that can do the very specific things that they need to have done and nobody else in the world needs except for them.
Is there something you don’t publish?
The people who were more committed stuck around and eventually became my cofounders for the company. The people who were more peripherally interested made some contributions and they went on to some other project or got busy with something else.
One is a software developer. The other is our customer support and salesperson. I am the CEO, do the hardware design and interface with the manufacturing partner.
They mostly find us. People will come onto the forum and they’ll post about something that they’re working on.
How did you end up working on this?
I could have published the paper without anybody interested in it, maybe only a couple hundred people read it. Then I would not have a team and much momentum. It could have ended there. But people were interested in it and it gave me confidence that this could be successful.
It’s about 3000 sold FarmBots up to now.
Being open source -what it really does-, it empowers the people who are passionate about the technology itself, to improve the technology.
How many people worked on the project?
What are the occupations of the others?
What’s your main role?
What are the occupation of the volunteer working people?
How do you find suitable people?