1 Year in Review, in Quarterly Updates
A little over a year ago I was just beginning to assemble the FarmBot team, was building the V1 hardware prototype, brainstorming what OpenFarm was slated to become, and newly hired at the SLO MakerSpace as the shop manager. While perusing the Open Source Ecology (OSE) wiki to see how FarmBot, OpenFarm, and OSE overlapped, I came across the logo for one of OSE’s funders: the Shuttleworth Foundation. Curious, I clicked the link. What I ended up finding turned out to be the perfect organization that I needed to help me follow my dream of working on FarmBot and OpenFarm full-time. They funded leaders at the forefront of social change and innovation, across all industries, with the one uniting factor being that openness is at the core of every project – perfect.
The application was brief: just a few pages of written responses and a 5 minute video. I happened to already be past the application deadline by a few hours, but proceeded to rush together my application that night and film and edit a video the next morning. I sent it in with no expectations. To my surprise, my application was still accepted late and I had made the ‘short-list’. After sending in some letters of recommendation and having a Skype interview, I waited some more.
On January 20th, 2014 I was notified that I had been offered a Shuttleworth Foundation Fellowship worth up to $125,000 in salary, project funds, and travel, along with a priceless amount of mentorship, fellows community support, and connections. I excitedly accepted the offer and was working full-time on the FarmBot and OpenFarm projects by April.
In the last year of being a Fellow, I feel that I have made tremendous progress on both FarmBot and OpenFarm, and I have grown both personally and professionally. As part of being a Fellow, I was asked to write brief quarterly reports of what I had accomplished, what I was talking about, where I traveled, etc. Here are my 4 reports from the year:
May 2014
What did you make?
- On March 31, I published V1 hardware design files on GitHub here. This was the first time any design files were published.
- Over the next few weeks, I designed and built V2 hardware. Here is a video of it moving for the first time. I published the design files on GitHub on April 20. I wrote a brief writeup on the changes from V1 to V2 on our wiki.
- After learning a lot from V2, I immediately began designing and building V3 which includes a lot of new features, most notably: endstops, which tell the machine when it is at the end of itself. This was constructed just before leaving for Budapest, and it was successfully connected to the Internet for developer access. See some photos here and the live stream of the bot.
- On May 31st, I gave a TEDx talk at UCLA. Video recording is not online yet.
- We have functioning software on the Arduino and Raspberry Pi that can receive a schedule of commands from the Internet and make the robot move!
- We have a basic functioning Decision Support System and a few test modules. This creates an optimized schedule of operations and is perhaps the most complex part of all of our software.
- We have a basic FarmBot web app with a staging server here. When logged in, we can connect with the bot and send it commands through a console within the browser.
Where did you go?
- Belgium to meet Tim Evers, one of the core software contributors. This was great to build team rapport and understanding, as well as talk at length about future plans for the project. It was convenient to go here on my way to Budapest.
- Washington, DC for an Agile conference where OpenFarm was selected to be the subject an agile development sprint for conference attendees. I took the opportunity to fly out Rick Carlino as well (another core software developer) to again build rapport.
What have you been talking about?
- Connecting FarmBot to the web, business models, financial sustainability, strategies for a successful crowdfunding launch, other ideas in other areas such as transportation and politics.
Who is joining your thinking?
- Rick and Tim have both become quite involved in the project. Tim is looking towards the future across all subjects with me, and Rick is excited about the crowdfunding launch of the FarmBot project and becoming financially sustainable.
What was your biggest win over the past 3 months – and what did you learn from it?
- Releasing early and often. By pushing out a release early, we learn quickly what is good and bad about the design. We can then make quick course adjustments to keep us moving in the right direction without losing a lot of time. It’s easy to stay at the computer for too long continuously trying to improve or optimize something, rather than just stopping and building it to see what really needs improvement and optimization.
What was your biggest loss over the past 3 month – and what did you learn from it?
- Documentation. When presented with the day’s tasks, documentation is always on there, but I never seem to get around to it because it is one of the least fun parts of the job for me. I need to work on this and just set aside a few hours a week for it and bite the bullet. I learned that without proper documentation, people will go down the wrong path and waste their time and money. This happened with Matt Lech as he began building a V2 bot without realizing the software was not yet functioning and the hardware was missing vital components like endstops and cable management.
What is the shape of your team?
(How many / Roles / Employee or Contractor)
- Me – Project Organizer, Lead Mechanical Engineer, UI/UX Designer
- Rick Carlino – Volunteer working on the web apps for FarmBot and OpenFarm
- Tim Evers – Volunteer working on the embedded (microcontroller and Raspberry Pi) softwares
- Matt Lech – Volunteer working with Tim and also building/tinkering with hardware
- Cameron Dunn and Torrin Smith – Volunteer students working on the Decision Support System (DSS)
- Scott Herrington, Bryan Rodriguez, and James Cruz – Volunteer students working on the universal tool mount system, water sprayer, and seed injector.
- Ghislaine Guerin – Volunteer working on the frontend of the FarmBot web app
What are the metrics you are tracking yourself against – and where are you?
- I wanted to release a new version of the hardware every 2 months. So far I am on track for that goal. I don’t put any metrics or deadlines on the volunteers, but they continue to exceed my expectations and become more and more committed to the project. The long-term goal has been to have a crowdfunding launch in Q1 of 2015, and I think we are still on a trajectory for that to happen. I need to work on figuring out better metrics.
August 2014
What did you make?
- My TEDxUCLA Talk about FarmBot is now online here.
- Launched a new wiki using MediaWiki software. See it at wiki.farmbot.cc
- Caught up on all of my documentation for the hardware development side
- Have developed a process for continuing documentation – Do it as I go, copy and paste stuff from past versions, take pictures during the whole process, etc
- Built V4 hardware with notable improvements over V3 being:
- Electronics enclosures/housings
- Cable carrier cable management
- Magnetic universal tool mount design
- Began designing V5 hardware with these improvements:
- Rotary encoders on motors
- Integrated water and vacuum into universal tool mount
- A plan for handling tools and their communication with the Arduino
- Took apart V1 hardware and shared the environmental damage here
- Updated codebases on the Arduino and Raspberry Pi now feature:
- Smooth acceleration of the machine’s movements (no longer as jerky)
- The use of g-code for all commands (faster movements and respect for the industry standard for CNC machines)
- Endstop, position, and error reporting
- Device selection and the beginning of the manual control widgets on the FarmBot web app
- Launched the OpenFarm Kickstarter campaign giving us:
- Over $21k in funding (with 1 week to go!)
- 1,300+ real-world beta testers
- A resurgence of development activity and a handful of new software and UI/UX contributors
- Started the process and conversation of incorporation
- Began publishing the project’s budget on our wiki here
Where did you go?
- Nowhere. But that was nice 🙂
- Applied to the Hackaday Prize which could take me to SPACE! hackaday.io/project/2552 Currently in the semifinals round (top 50)
What have you been talking about?
- Incorporation of FarmBot and equity stakes
- How to split up equity moving forward, open value networks
- How to run a business openly (publishing budgets, public salary calculators, public board minutes, etc)
- Developing partnerships between OpenFarm and universities/other organizations to help build OpenFarm’s content
- Developing a community around OpenFarm, not just building another database
- Re-imagining/re-inventing all of government and the political process with modern trends and technologies. (Very excited to learn more about Seamus and the Madison Project!)
Who is joining your thinking?
- Rick and Tim have been part of the incorporation/equity discussion. I feel really good about them and we are all on the same page. Looking at giving each of them 10%.
- Folks from other Open Source Projects (OSE, OSBH, Sensorica, ) have been contacting me and we’ve been in loose talks about future collaboration, open business models, technologies/platforms for sharing content and engaging the community, the future of open tech…
- The whole OpenFarm dev team is very involved in discussing the community aspects and partnerships
What was your biggest win over the past 3 months – and what did you learn from it?
- Launching [the OpenFarm Kickstarter](http://kck.st/1yBHkVG) proved to be extremely valuable for more than just the money:
- The volunteer development team tripled in size! There are now 13 contributors, 44 watchers, and 35 forks!
- We now have over 1,500 beta testers (the backers), many of whom have already made accounts on OpenFarm
- Running the Kickstarter was great practice for the FarmBot Kickstarter coming up next year. Will be making a writeup with all of the things learned through the process, and a plan for the FarmBot campaign including:
- Make solid connections/relationships with bloggers and news writers well ahead of launching
- Have interview questions, message replies, and comment replies pre-written in a spreadsheet and ready to go
- Have press releases and blog posts ready at launch
- Basically, do as much work as you can before the launch happens
- Be a whore on Twitter, Facebook, and Google+
- Don’t blow it on Reddit – be a part of the subs before posting, make sure to not get auto-moderated off
- Always share the the short link from Kickstarter and then check out traffic sources on bit.ly (the service Kickstarter uses)
- Get the word out to the right people, not the most people
- Most folks will come to you from news sites, not from your own postings
- Many more things!
- And of course, we did raise over $22k!
What was your biggest loss over the past 3 month – and what did you learn from it?
- I burnt myself out at the end of June. I was working on too many things, for too many hours, and not taking enough time to myself. When so excited about the project, it can be easy to focus too much on it and let other factors of life fall by the wayside: sleep, eating well, exercise, etc. Working at such a high level in the time after the Gathering was unsustainable and came to bite me after about a month.
- I now recognize more clearly the importance of taking care of myself in order to maintain a high level of work
- I have found a new balance between work and personal time that allows me to enjoy each one more than when one is put on the backburner.
- I have been learning to prioritize, delegate, and focus more effectively to move the project forward the fastest and most efficiently.
What is the shape of your team?
(How many / Roles / Employee or Contractor)
- Me – Project Organizer, Lead Mechanical Engineer, UI/UX Designer
- Rick Carlino – Volunteer working on the web apps for FarmBot and OpenFarm, will have ~10% equity in FarmBot Inc soon
- Tim Evers – Volunteer working on the embedded (microcontroller and Raspberry Pi) softwares, will have ~10% equity in FarmBot Inc soon
- Matt Lech – Volunteer working with Tim and also building/tinkering with hardware
- Ryan McLeod – Volunteer working on OpenFarm frontend – candidate for OpenFarm contract work
- Jeremy Kauffman – Volunteer making mockups and helping with UX for OpenFarm
- Scott Herrington, Bryan Rodriguez, and James Cruz – Volunteer students working on the universal tool mount system, water sprayer, and seed injector
- Simon Vansintjan – Baller volunteer working on all aspects of OpenFarm – candidate for becoming the OpenFarm ‘Champion’
- Richard Leland – Volunteer for OpenFarm
- A handful of other occasional contributors
- Have started the process of getting more students/student teams on board from the local university (soil scientists, mechanical engineers, software developers) School starts at mid-September
What are the metrics you are tracking yourself against – and where are you?
- Still trying to get out FarmBot releases early and often (every 2 months). In the 6 months since starting the fellowship, I’ve released three versions (V2 thru 4) and V5 is in the pipeline. So I am still on track with this metric.
- I don’t know what the metric would be called, maybe team morale/rapport/commitment/excitement, or how to track such a metric. But it has been really important for me to build the volunteer team, especially with Rick and Tim who are going to become equity holders. Again, I don’t know how to track it other than with a gut feeling of how things are going, but it seems to be going quite well! They are still very excited to work on the project, and feel as dedicated as ever 🙂
November 2014
What did you make?
- Launched a new, modern, sexy FarmBot Website in order to ‘sell’ the idea more legitimately during the Kickstarter campaign. See it here.
- Launched a new, more modern forum for FarmBot. Very new and not much activity yet, but it is here.
- Built V5 hardware, tested it, documented it on the wiki here.
- Built V6 hardware, tested it, setup a live stream feed for our developers and the rest of the world here.
- V7 hardware design is nearly complete.
- First semi-working versions of the seed injector, watering nozzle, and liquid amendment tools have been built and tested by the mechanical team, still need to document the work on the wiki.
- After much feedback and several iterations, finalized the design the OpenFarm Growing Guide Creation Flow to support structured data entry in a fun and easy way. Mockups are here. The beginnings of it can be seen on staging.openfarm.cc.
- Rethought what search should look like on OpenFarm, designed mockups, built it, its on staging.
- Developed a Code of Conduct for OpenFarm – will be applied to FarmBot too.
- Feature released manual controls on the FarmBot web app for the vacuum pump, watering valve, LED, and Tool Mount
- Began high level designing how users will create sequences of commands for FarmBot and then schedule them in bulk.
- New OpenFarm homepage design and implementation. See it on production.
- We imported 40k crop scientific and common names into OpenFarm. We have another 90k to cleanup up a bit and import.
Where did you go?
- I visited Daniel Lombrana in Madrid to give a talk about my work at the Laboratorio Urbano Workshop in Madrid.
- After Madrid I went to the North West of Spain to an off-grid farm to stay with Andru Vallance for a few days. He is the creator of Practical Plants and a strong ally in the space of open plant/agricultural data. We had some epic brainstorming sessions about the future of OpenFarm.
- Malta for the Gathering!
What have you been talking about?
- Been extra active on the OpenFarm Blog talking about the design and data structure, we’re trying to post every week or two.
- I gave a talk about projects that matter at the Madrid workshop.
- Making OpenFarm a non-profit sooner rather than later in order to get more money and attract partners.
Who is joining your thinking?
- I got really great in-person feedback on my Madrid talk and what I am doing with my work. That is always really rewarding to hear!
- The OpenFarm team and I had a lengthy discussion about the future of OpenFarm and what our Mission and Vision are. I was the strongest voice in the conversation and I felt that the other team members were joining me in my vision. Great team building discussion and good to get us all on the same page moving forward!
- The OpenFarm community at-large via our public discussion group on Loomio.
- The OpenFarm business/entrepreneurship students with regards to becoming a non-profit
What was your biggest win over the past 3 months – and what did you learn from it?
- There have been two new FarmBot replications documented on the wiki! One here and the other here. And one more just a few hundred miles from me that I have seen progress on via direct emails!
- I’m learning the value of open documentation, how it inspires others, guides them in their design, prevents duplicate work, and how it is is a feedback mechanism for me to learn more as well.
What was your biggest loss over the past 3 month – and what did you learn from it?
- Letting FarmBot stuff go a little to the wayside because of OpenFarm development.
- This set us behind schedule for where we wanted to be with FarmBot by now (still fighting with the manual control page, no bot outside of the lab environment yet)
- We’re realizing the monumental task that it is to create the operation scheduling interface on the web app
- However, with the long-term vision in mind, OpenFarm and its data will be a vital component to FarmBot’s success and more automated abilities, and creating a robust dataset on OpenFarm is going to take time too, lots of time.
What is the shape of your team?
(How many / Roles / Employee or Contractor)
- Me – Project Organizer, Lead Mechanical Engineer, UI/UX Designer
- Rick Carlino – Now paid part-time using my project funds to work on the FarmBot Web App
- Tim Evers – Volunteer working on the embedded (microcontroller and Raspberry Pi) softwares
- Matt Lech – Volunteer working with Tim and also building/tinkering with hardware
- Ryan McLeod – Frontend dev doing contract work for OpenFarm
- Gabriel Burnworth – Volunteer helping with FarmBot miscellaneous stuff (building prototypes, design review, scripting)
- Kevin Bertolero and Katherine Ying – Contracted as Community Managers for OpenFarm, peripherally helping out now, will start officially with the launch of OpenFarm in January
- Jeremy Kauffman – Volunteer making mockups and helping with UX for OpenFarm
- Scott Herrington, Bryan Rodriguez, and James Cruz – Volunteer students working on the universal tool mount system, water sprayer, and seed injector
- Simon Vansintjan – Baller volunteer working on all aspects of OpenFarm (backend, frontend, visioning, blogging, outreach, etc)
- Allyson Woods, Jake Wing, Christian Renfro, Emily Plummer are business entrepreneurship students working on OpenFarm customer development, becoming a non-profit, and answering the question: what happens when the Kickstarter money runs out?
- A handful of other occasional contributors
What are the metrics you are tracking yourself against – and where are you?
- We’re now using Google Analytics to track everything on OpenFarm.
- Uniques per month: 4.1k
- Average Session Duration: 1:59
- For FarmBot, still pushing out a new version every 2 months. Released V5 and 6 since last quarterly update, still on track and making significant improvements with each version
- Will now start tracking how many independent replications are happening, currently at 4.
February 2015
What did you make?
- Cloned and reformatted a lot of wiki documentation onto the FarmBot website to begin strengthening my brand and separating my to-be company’s work from the community’s:[1]
- Finished FarmBot V7 hardware design and prototype. This is the first functional “large” prototype after V1. [2]
- Further iterated the Universal Tool Mounting System, watering tool, and seed injector.
- Launched the OpenFarm website on January 18th with support for Guide creation, user gardens, and an API matched to our feature set [3]
- Learned how to program(!!) and styled the FarmBot web app: [4]
- Sequence building and sequence scheduling are now minimally functional on the web app.
- Began working with Content Mine to mine scientific plant facts for OpenFarm
- OpenFarm’s mission and vision statements: [5]
Where did you go?
- North Carolina for a talk at the Produce Marketing Association’s ‘Tech Talks’ event [6]
- Chicago for ORD Camp and to meet with Rick [7]
What have you been talking about?
- Open source affecting every part of our society: [8]
- Open source is going to disrupt the ag industry [9]
- Launch strategy and plans for FarmBot: [10]
- Web app roadmap: [11]
- Manual Control page design [12]
Who is joining your thinking?
- I got really great in-person and Twitter feedback on my talk at the Produce Marketing Association’s conference [13]
- Folks who join my teams and my efforts are joining my thinking. For example, people who have started their own FarmBot Projects on the wiki, and those who have joined our OpenFarm Slack group
What was your biggest win over the past 3 months – and what did you learn from it?
- We launched OpenFarm!
- I learned the value of shipping early and often. By anyone’s *normal* standards, OpenFarm was not ready for launch. But I pulled the trigger anyways and it was successful. Wegot tons of feedback, press, new contributors, and a new fire was lit up under everyone. If I hadn’t set the date and pulled the trigger, we probably still wouldn’t have been launched by now!
- Learning to do frontend engineering. Amazing how empowered I am with this extra skill!
What was your biggest loss over the past 3 month – and what did you learn from it?
- Not having got the ball rolling on developing the FarmBot web application sooner, especially in regards to having the project pitch approved.
- I didn’t expect the process to have a contract written up for Rick to take so long, or for my project pitch to be approved.
- Because of this, I feel like the web app is 2 months behind where I would have hoped it would be.
- I really should have pushed the project pitch through about 3 months into my fellowship rather than 9 months in. I guess I was being conservative and careful with the project funds, and I should have been more aggressive and just opened the flood gates to speed things up.
What is the shape of your team?
(How many / Roles / Employee or Contractor)
- Me – Project Organizer, Lead Mechanical Engineer, UI/UX Designer, Frontend Engineer
- Rick Carlino – Almost being paid using my project funds to work on the FarmBot Web App
- Tim Evers – Volunteer working on the embedded (microcontroller and Raspberry Pi) softwares
- Matt Lech – Volunteer working with Tim and also building/tinkering with hardware
- Ryan McLeod – Frontend dev doing contract work for OpenFarm
- Gabriel Burnworth – Volunteer helping with FarmBot miscellaneous stuff (building prototypes, design review, scripting)
- Kevin Bertolero and Katherine Ying – Contracted as Community Developers for OpenFarm
- Jeremy Kauffman – Volunteer making mockups and helping with UX for OpenFarm
- Simon Vansintjan – Baller volunteer working on all aspects of OpenFarm (backend, frontend, visioning, blogging, outreach, etc) Now being paid for a small contract
- Allyson Woods, Jake Wing, Christian Renfro, Emily Plummer are business entrepreneurship students working on OpenFarm customer development, becoming a non-profit, and answering the question: what happens when the Kickstarter money runs out?
- Rich McClusky – Frontend Engineering for OpenFarm
- Claire Dierksen – Graphic design and brand development for OpenFarm
- Matt Pruitt – Dev ops for OpenFarm
- Jake Rayson – Frontend engineering for OpenFarm
- A handful of other occasional contributors
What are the metrics you are tracking yourself against – and where are you?
- OpenFarm
- Uniques per month: 6k (up from 4.1k last update)
- Average Session Duration: 3:16 (up from 1:59 from last quarterly update)
- Registered users: 2,277
- FarmBot
- Still pushing out a new version every 2 months. Biult V7 since last quarterly update
- On track for a Kickstarter in August/September
- Will now start tracking how many independent replications are happening, currently at 6.
