The two most common types of constructive criticism we’ve received from the beta launch of AutoNanoFarm are about the price and the aesthetics. It’s an interesting pair of criticisms, because at low volumes, these two factors compete with each other. Reusing off-the-shelf components that are already manufactured at high volumes for other purposes helps keep the cost down, but doesn’t allow for the most attractive design. At high volumes however, a custom design for AutoNanoFarm can be inexpensive.
We have the nagging feeling that forcing off-the-shelf components such as an aquarium, a custom stand, and plastic containers as vegetable beds together into an elegant system is not the best way to go about it. There has to be a better way to create a system that accomplishes the same goals, but in a sleeker and more elegant package. After all, aside from its practical value as a small garden, and its educational value as an introduction to aquaponics, AutoNanoFarm is also meant to take a spot in your home as a conversation piece.
This thought would not leave us alone, so finally, we decided to do something about it. We were inspired by the elegance ofJellyFishArt aquariums, and wanted to create something similarly sleek. Also, the nautilus spiral kept returning to mind, and this is what we came up with:
In our preliminary design, the tank would hold about 20 gallons of water when reasonably full. the pump would be hidden in the top wooden spiral, and a misting system (aeroponics, to be technical) would deliver the water to the plants before returning back to the aquarium.
What are your thoughts on this idea? Here are a few tradeoffs compared to the current AutoNanoFarm:
Let us know what you think via your reply. As always, we look forward to hearing from you!
Finally, the day we’ve long been working towards has arrived: AutoNanoFarm is ready for sale!
Let us re-introduce AutoNanoFarm to you. AutoNanoFarm is a small aquaponics system that easily fits in your kitchen. This is a good system to start out with, whether you live in an apartment or you are looking into starting aquaponics and want to start with a small system before moving to larger systems. By getting familiar with a smaller system such as the AutoNanoFarm, you will be able to transfer the experience to the larger scale AutoMicroFarm that is in the development stages.
The AutoNanoFarm system uses a custom-built wood cabinet to hold a 20+ gallon fish tank and a container with two vegetable bed inserts on the top. The system utilizes a flood and drain method to pump the fish-fertilized water from the fish tank up to the growing beds at regular intervals, and drains back down. The growing beds filter the water and use the nutrients from the fish water for growing veggies, fruits, herbs, etc… whatever you plant. You can read more detail about the AutoNanoFarm here, and see pictures of it during assembly here.
Your AutoNanoFarm garden will be watered automatically. You just need to feed your fish, and add some water weekly to keep the tank full—you don’t have to change out the water, since the fish waste in the water will be used as fertilizer by the plants.
We have the kinks worked out, and we have a functioning AutoNanoFarm that has been running well. Below are actual pictures of the product, not renderings.
Front View of the AutoNanoFarm
The Growing Beds
The Aquarium, with two hard-to-see bluegill that we’ve taken from our first prototype.
Please note that this product is truly a beta product: there are things that you may find that are not ideal (for example, the flood-and-drain system we are currently using will make your aquarium water level rise and fall by about 25-30%. We have ideas on how to keep the water level fluctuations to a minimum, but we’re still experimenting with them). This is why we want your comments and opinions for further development based on your feedback. Multiple stain colors are available to stain the wood frame such as dark cherry, walnut, etc. The vegetable containers will be slightly different, but still using food safe plastics.
The beta version is ideal for you if:
If this is you, we would like to offer you an AutoNanoFarm at cost: $400. We will discuss the color you prefer (see below for the choices), and after a few weeks, deliver one to you personally, and set it up at your house. It will include the stand, the vegetable beds, the aquarium, the pump, and the plumbing. You will just need to provide the fish and the plants/seeds (or we can provide them, if you prefer, also at cost). Please reply to this email to order your AutoNanoFarm. If you would like to see our functional prototype in person, please get in touch with us to arrange it.
We’re eager to hear ideas and suggestions for improvement from you, this first group of early customers!
Michael of RestBackup.com recently wrote us AutoMicroFarmers a note of encouragement:
Thanks for the update. I think your project is really cool. I hope you can make it into a profitable business before running out of money like I did [with a previous project].
We were thrilled to receive the encouragement (and great advice in the rest of his email). After reflecting on his words, this is what we wrote back:
We’re hoping to become profitable within a reasonable timeframe, as well, but our financial runway is infinite: we’re bootstrapping, but more than that, we’re spending such a low amount (for a hardware project) that finances are not an issue.
Our biggest scarce resource, believe it or not, is interest/ambition/passion/enthusiasm, which we did not expect to be limited. We’re looking for ways to increase that, so any tips on strategies would be great and appreciated.
When we started analyzing and discussing possible reasons why our passion might be waning, we came up with several.
First of all, startups are hard: there are many setbacks, and you have to border on being naively, foolishly optimistic that you will succeed against the odds. That’s hard to do for smart, realistic people who are prone to cynicism.
Also, estimating progress, while having many benefits, also has downsides: what we thought should have been a trivial problem with a no-brainer solution actually took many hours of searching for the solution that will actually work. The slip-ups in estimation schedule are demoralizing. To all of you who thought we’d be further ahead of our development by now: we’re sorry. We’ve learned our lesson, and are more determined than ever not to make (or imply) promises we cannot keep. (It’ll be ready when it’s ready!)
Still, that leaves us with a question: how do we keep keep our interest and passion from waning?
One thing we’ve decided early on is to put a minimum and a maximum weekly number of hours we work on AutoMicroFarm: the maximum to make sure we keep a balance on our lives and don’t burn out (it is a side project, after all), and the minimum to make sure we make progress every week. We’ve gone below the minimum on a few weeks, and really regretted it: things got better, and we realized we’d be further ahead if we had just pushed through.
We also found good advice in several articles around the web:
The 37Signals blog has some good advice on this: ignore details early on, the middle is what’s important, seek and celebrate small victories.
We also got really good advice on Sebastian Marshall’s blog when we asked his readers the same question; since then, Sebastian posted a really good post on seeing the benefits right away when working on a project.
Finally, Lukas Fittl wrote an excellent post on when to come up with new ideas, and when to execute on them as a guest on Ash Maurya’s blog. Mixing the ideation and execution modes willy-nilly can have a big negative impact on the morale and passion of a startup team.
In conclusion, the best way to increase our passion is to be successful as a startup… now we just have to figure out how to do that :). One thing we’re pretty sure of, our passion will get a good boost when an actual customer deposits money for the first product. Speaking of which, look for pictures of a fully functioning AutoNanoFarm very soon (but we won’t fall for actually promising a date when they’ll be posted!)
Do you have any (other) advice for us? As always, we’d love to hear from you.
Hi Everyone,
We are sorry we’ve gone so long without an update. Now that we’ve finally gotten around to writing it, we want to tell you about two things.
Progress on our project has been like riding a roller coaster.
We get a lot accomplished one month, and the next—almost nothing. But we do have some good news. We finally have a commercial prototype of the AutoNanoFarm stand designed and custom built by a custom cabinet maker also located within the Research Triangle area of North Carolina, where we live. (We’ve posted pictures of the components here.)
The container and inserts are off the shelf parts from Target. We really didn’t want to spend the time trying to weld a custom plastic container and inserts, or fiberglass one, so we decided to give a couple local stores a second try. At first, we had no luck finding any inserts that had a square footprint and a container that was able to fit several inserts into it. Then it finally struck us: why “square”? So we picked up the first rectangular “inserts” on the shelf next to us and they fit perfectly into a larger container.
Then came the drilling of holes for water. We learned a good lesson for $7. If you are going to drill a hole into a clear plastic container, have the drill spin backwards. The first drill hole shattered the container and a ¼” hole turned into an ugly web cracked 1” hole. The crystallized structure of a clear container makes actual drilling a pain. The plastic can’t take the shear force and cracks right along the formation. The next hole we tried to drill in reverse by literally heating up the plastic and creating an opening without applying shear stress. IT WORKED! Forty holes later, and we have our inserts.
When we started to put it all together, we ran into several different problems that we’re still shaking out. It turns out that neither silicone nor other glues work well on rubber and plastic for a water-tight seal. However, just yesterday, we finally succeeded in getting a leak-proof system by ditching the rubber uniseal and using JB WaterWeld on the plastic parts. Now that we have an operational system, we’ll test it out to see if any minor adjustments and improvements have to be made, before we start making our first ten units for local aquaponics enthusiasts (if you are in the Triangle area and are interested in being a beta-tester, please let us know). We are already thinking of re-working the height of the AutoNanoFarm, because the veggie bed currently sits a bit too high (44.5” above the floor). Lower the top 4.5” will make the AutoNanoFarm fit right below the window without blocking the view outside.
We will have multiple stain and paint color options available along with different quality woods to choose from. Cherry, Oak, Maple, and others will make it a furniture piece that fits in perfectly with your current home furniture and decor.
We are also in the process of updating our website, so it is more user friendly and easier to navigate (including a newsfeed to our blog on the main page).
We are hoping to get some feedback and involvement from you.
We need your feedback on renaming AutoNanoFarm to a more commercially friendly name. AutoNanoFarm is just a working product name: it might be too geeky for the general population. Do you have any suggestions for the product name? We’ll take the most promising ones and create a poll on our blog.
We also need an off-the-shelf 24” or 36” x 12” x 12” inner dimension plastic container, if we can’t find one, we’ll custom manufacture it, but we would rather not. If you know where we could get such a container for a reasonable price ($10-$25 each), please comment with a link!
[video]
Happy New Year everyone!
We wanted to give you an update on our AutoMicroFarm progress. We’re making good progress on the AutoNanoFarm, with the design mostly finalized and the prototype well under way.
We have a vegetable bed and two inserts acquired as off-the-shelf plastic containers (we’re planning on drilling out holes in the inserts for our first couple prototypes). We are working with a local furniture manufacturer to make a high-quality wood stand to support the veggie bed and the aquarium. We are also going to use a standard 24” L x 12” W x 16” H 20-gallon aquarium that matches well with the 16.5 gallon-vegetable bed, assuming you fill the aquarium 75% full.
We will also start emailing this message to our 1500-strong email list. If, as a venerable twitter user, you want to receive these emails as well, feel free to sign up with your email at AutoMicroFarm.com. If you haven’t already, follow us on various social networks and share our posts with friends.
As always, feel free to get in touch with us (just email info at our domain), we love chatting with people interested in AutoMicroFarm.
We AutoMicroFarmers have gone a month without any updates. However, we had a rather significant event occur that we could not talk about until now.
The fact is, we got invited to Y Combinator for an interview, and spent the two weeks leading up to the interview preparing as much as we could. (Special thanks to AnyVivo, ViaCycle, and Double Robotics for giving us advice when we cold-emailed them.) We thought the interview, which occurred on December 1st, went really well; we didn’t really have any questions catch us off-guard.
However, a few hours after the interview, we got the following email from a YC partner:
I’m sorry to say we decided not to fund you. We liked you guys and were impressed with your ambitious ideas for aquaponics. In the end, it was difficult for us to see how this product would generate the kind of growth startup investors are looking for. We would not want want to dissuade you from pursuing the idea however, and think something like this could be quite effective on a crowdfunding platform like Kickstarter.
We can see where YC is coming from. After all, there are currently several tens of thousands of aquapons in the US [1], and in the near future, we hope to be able to sell that number of AutoMicroFarm products annually. Thus, we hope to single-handedly, as it were, expand the aquaponics market by offering convenient, inexpensive, easy-to-use, and eventually automated aquaponics systems. Perhaps that’s a bit too naive and ambitious…
However, we genuinely believe that if we can accomplish our objectives and offer systems that are both high-quality and inexpensive, every kitchen would benefit from an AutoNanoFarm, and every back yard from an AutoMicroFarm. That’s what we want to see: an AutoNanoFarm in every kitchen, and an AutoMicroFarm in every back yard [2].
We definitely plan to run a kickstarter campaign in the near future. However, before that occurs, we would like to custom-build several (perhaps up to ten) AutoNanoFarms, and have interested customers in the Triangle, NC area buy them. To those customers, we offer free and unlimited support and troubleshooting, in exchange for extensive feedback about our product. The things we will learn from such an initial effort will help us achieve the high quality that we need.
Please let us know if you’re interested in being a trial user of the AutoNanoFarm, and you reside in the triangle. We would love to work with you!
1. This is our rough estimate based on several facts we were able to find online. If anyone has a more accurate or precise number, please let us know.
2. Or at least, in every third kitchen and back yard :).
TL;DR: To make maintaining and changing things in an AutoMicroFarm easier and more convenient, we’ve come up with a concept of inserts. We need your suggestions on how to best prototype and manufacture them. We would love to hear from you!
Although aquaponic systems are very easy to use to compared to gardening, one of the few things we found to be the really inconvenient from our barrel-ponics prototype is maintaining the vegetable bed (or containers). For example, tomato plants have very extensive root systems that tend to dominate the media bed, and start growing into the siphon/outlet area (something we found out first-hand). Also, we’re sure many of the first AutoMicroFarm customers will want to experiment with different setups and conditions. For this reason, we want to make sure that our aquaponics systems are as easy to use, maintain, re-configure and change as possible.
The solution that we’ve come up with both on our own, and in discussion with other aquaponics experts, is to use inserts. Think planting baskets used with aquatic plants (Figure 1 below), but uniquely suited for an aquaponic system.
Figure 2 below depicts the side view of the inserts, each consisting of a support structure and a space for the media. We have examples of very little media, on the left; progressively less media space in the middle; and where all of the space is taken by the media, on the right (the example on the right is functionally identical to the planting basket above, but bigger). Of course, more space configurations are possible. In all cases, the insert walls will have holes to let the roots grow out of the media into the open space; the holes will be small enough to prevent the media from falling out. Also, the tops of the containers will touch to prevent light from reaching the plant roots, which causes undesirable changes in most plants.
The bottom half of Figure 2 above shows the same three inserts in a container, with the middle inserts taken out. Each insert will have a support structure strong enough to hold the media and the plants, and handles (not shown in the figures above). We’re in the process of designing these inserts. The overall dimensions will tentatively be 12” H x 16” W x 16” L (30cm H x 40cm W x 40cm L). We have also considered smaller widths and lengths, including 12” (30cm) and 8” (20cm). Once we start with dwarf/pseudo-bonsai fruit trees, we may also have bigger inserts than 16”x16”. All these design considerations will be tested and iterated on to find out what works best.
We think there is real value added by incorporating inserts in AutoMicroFarm systems, instead of just filling the container full of media—and not just for the tinkerers. By using inserts in our next product, we will be able to do maintenance on the root system easily by being able to take out an insert and get to the bottom of the root system, doing any trimming if needed. Also, the inherent compartmentalization of each insert will also allow for quick re-planting of new vegetables, fruits and herbs in isolation from the rest of the vegetable bed for testing the best combinations of companion plants. Another benefit of the inserts that we recently realized is the ability to grow root vegetables, easily seeing how the root is doing, and harvesting it if needed.
In the next post on the topic of inserts, we’ll share our work on designing and researching manufacturing options for the inserts and the vegetable bed, and the timeline for doing so.
For now, we have several requests: please suggest any improvements you can think of from the stand-point of convenience and ease of use. If you have aquaponic systems and have any concerns, let us know. Also, if you have experience manufacturing these sorts of things, please give us suggestions on how we can best prototype and manufacture these inserts. We would love to hear from you and collaborate with you on making AutoMicroFarm systems the world’s best aquaponic systems!
TL;DR: We are winding down the barrel-ponics proof-of-concept experiment. We achieved 50% of the harvest yield that we had hoped for, in the bountiful periods. But there are a lot of improvements we’ll implement in our second prototype, which will help us come closer to our target.
Below is a graph of yield per week over roughly the last six months. We averaged 50% of what we calculated to be the ideal yield per unit area during the better periods, and about 35% overall.
Although we did not reach our goal, there were tons of improvements we could have done to increase the yield. First of all, we did not harvest any fish during this time period. We also made the typical rookie mistakes of not waiting for the initial cycle to end, and not measuring the ammonia/nitrite/nitrate levels in the beginning to see what was going on, as mentioned in an earlier blog post. Also, most of the time, we had a lot of empty patches where we could have planted other vegetables. So, in short, our first prototype was nowhere near optimal, and thus we did not see optimal yields.
Another discovery that we made during this time is that our assumption of the amount of food needed to make up 90% of an adult’s diet were way off. Our assumptions might hold for vegetarians or vegans who eat a lot of plant-based foods, but even our family, which probably eats more fruits, vegetables, beans and nuts than the average family in the US, would get all the food we needed from just 50 to 100 square feet (the area we thought 1 to 2 adults would need for 90% of their diet).
Our next steps are to make the next prototype, with the focus on making the system more convenient to change when you want to try something else, or fix when things go wrong. Incidentally, this prototype will be our first AutoNanoFarm prototype. We have some exciting ideas we’re trying out right now and designing into the product, so stay tuned!
This past Wednesday, we got a spike of visitors to the AutoMicroFarm site, and not only that, but over a thousand 1200 people signed up to our mailing list since then. So what is all the excitement about and where did it come from? On Wednesday morning, one of us made an off-hand comment in a post on Hacker News about where AutoMicroFarm ranks on the Yudkowsky Ambition Scale. Perhaps it helped that Mr. Yudkowsky chimed in with appreciation. Within minutes, we had a spike of visitors to our site, including messages about how excited the new visitors are about our product and what we are doing.
As a response to all of the traffic, we are updating our website for easier access to information about AutoMicroFarm. We are very excited about the enthusiasm of the subscribers to AutoMicroFarm. We are posting up a quick one-minute survey for you to take, so we have a little more information on where to focus our energy.
We are very close to releasing our initial design for the AutoNanoFarm, our first product (note the Nano, not Micro). For those that are into manufacturing, we are eager to hear from you with advice on our design (especially the manufacturability of the injection-molded parts). As we start solidifying the first iteration of the design, we would love to contact potential manufacturing partners to create both the prototypes, and eventually the mass production of the parts.
Thanks again, and keep in touch!
The AutoMicroFarmers