Homebrewing is a test. The reward: Beer whenever you want and your own unique recipe (plus beaucoup indie cred points). And this is no half-hearted attempt. Unlike pickling or infusions (other gastronomical pursuits du jour), homebrewing requires an investment of weeks, sometimes months, with a learning curve for each batch and style. But let's be real—no one wants to wait 4–6 weeks dreaming about how their beer is going to taste while it goes through some scientific fermentation process. Well, maybe you do. And that's ok, too.
But for those of you who would rather have your brew quicker, Kickstarter has an option for you. Microsoft developer Bill Mitchell teamed up with his brother Jim and friend Avi Geiger to launch an PicoBrew Zymatic.
At a glance, it seems like the contraption has basically cut out any risk of user error... but don't hold us to that.
At press time, PicoBrew's is nearing three times their funding goal of $150,000. Co-founder Avi Geiger gave us some insight into the creativity and process behind the project:
Co-founder Avi Geiger
Core77: How did you come up with the technology behind this and why do you think it hasn't been done before?
Avi Geiger: It comes down to having the right mix of people and the right mix of frustrations with the world. Usually when people try to automate brewing, they build complicated systems with multiple tanks, valves, little cranes and robotic arms to add hops. There are literally hundreds of homebuilt "automated" brewing systems—each one is different, most of them are very expensive. If one of these was the right answer for more people, then you would be able to buy it somewhere by now. Bill and Jim came at it from the product development and process side and developed a new system that allows for preloaded ingredients and automated fluid distribution without any valves. I came at it from the engineering side and collectively we worked together to get this into a compact and reliable package that takes this method and lets us brew beers to the quality standards of a professional brewer with the ease of use of a kitchen appliance.
Another reason it hasn't been done before is that it's just a ton of work to get all the details right and bring it to market. We've been working on this full time for over two years now. And there was a year of concept work before that. We've done hundreds of test batches to optimize the process. There are an incredible number of variables in beer production and picking the right set to work across was a pretty big step in itself.
Where was the "a-ha!" moment in the process?
There are so many I don't know where to start. We learn something new every day and only sometimes it's what we set out for. Some of the biggest things that let us move forward were figuring out which variables were most important in the process and how to control them.
How does your beer compare to conventional homebrew?
Quite well! In fact, just last week Annie Johnson, the American Homebrewers Association Brewer of the Year, visited the office and tasted them. Conveniently, she's also a beer judge for homebrew competitions. Of the 15 beers she tried, she thought that four or five of them could have won their category in a competition. That's pretty high praise. Our lab techs quickly bottled up a few of the beers and submitted them to a local competition and we're all excited to see how they do.
What's the technology behind your brewing process?
Brewing can be described pretty simply. Extracting sugar from the grain; sterilization in the boil; extracting bitterness, flavor and aroma from the hops. The fermentation to convert the sugars into alcohol. Unfortunately it usually involves large pots and open kettles, with constant attention over a long period of time if you want credible results. They call it a "brew day" for a reason. And of course the real glory is in the details.
Our core technology is about redeveloping that process into a countertop form factor that still gives professional results. There are a few common problems we wanted to solve outright like sterilization, which can be really difficult for brewers and their fermenters. It's a recirculating system so it's self-sterilizing during the boil process. By using a stainless keg as both our brew and fermentation vessel, it's had steam-sterilized boiling water circulated through it for this same process. We developed the control systems for consistency and tested them across hundreds of batches.
After that is compactness and ease of use. There are some real innovations there. One big difference is that we bring the fluid to the ingredients instead of the ingredients to the fluid. It's more complicated that it sounds when it comes to designing a manufacturable vessel that allows this in a compact space without adding time to the brew process—For example, the way we handle the hops or other adjuncts. Hops are typically added at various points through the boil process, so we've devised a system where the hops share a divided compartment that allow the fluid to waterfall through each hops addition. Cleanup is also a big deal. Few people enjoy scrubbing caramelized bits out of 10-gallon kettles with their garden hose. We designed it for the machine to be self-cleaning and the removable parts to be washed in the dishwasher.
Even with all the mechanical and electrical engineering we did, what really stands out is the software. The software takes everything we know about how this system can be characterized and translates that into tools for developing and sharing recipes and continuous improvement with every batch brewed.
And as far as we've taken this, it's only a start. There are tons of new features we just don't have the time to implement and many more that nobody has thought of yet. Brewers are a crafty bunch and we want to enable them to take this wherever they want to go. The firmware is open source—we'll have open APIs to our cloud-based services—and we'll make schematics available to those who ask.
What was the prototyping process like?
Prototyping has given us a ton of interesting challenges. It rarely helps if we can't brew on it, so everything we do has to be made from high temperature food safe materials, and those don't just pop out of 3D printers.
For example our plastics. Everything is polycarbonate, which doesn't fabricate well. It burns if you try to laser it, it's flexible enough to wreak havoc on vacuum hold downs and doesn't bond well to itself. Long-term, of course these parts will be injection molded but this is a huge tool and the costs and lead times don't allow for mistakes. We tried having local plastics fabricators take the job but they all said it was impossible so we had to develop an in-house process for it.
Similarly with the containers to hold our hops. In production these will have stainless mesh inserts molded into plastic, but that's not exactly a prototype process either. With four compartments per machine, five machines on our lab bench and six sides per compartment—and repeating this with every design iteration—we needed a more efficient process than a hot air gun. So we CNC'd our own hot stamping tools, rigged it up with cartridge heaters and a PID controller and mounted it on the arbor press for alignment. It was an afternoon project that probably saved us weeks of manual labor.
Did you have any entirely terrible batches along the way?
Why does everyone want to know about bad batches? I take it as a sign that the current batches are so good it's hard to imagine otherwise. "Bad batch" is relative. As long as you get a sterile sugar solution out of the machine, add yeast and ferment correctly you are going to get beer. (Conventional homebrew fails on the sterile part more than most people would admit.) If you don't contaminate it while taking samples or transferring from one keg to another, it's probably going to be more than drinkable. But what we're after is the same beer every time, per the recipe, with across a range of styles from a light lager to the darkest imperial stout.
We took a scientific approach to optimizing which means running huge grids of experiments. And every data point is a batch of beer. Characterizing mash extraction is relatively straightforward because you can take easy measurements during the process. Hops are a lot tricker. You have to wait until the batch ferments, then send it off for analysis. This process is about two weeks so we were ready to do anything we could to short-circuit this. We tried all kinds of experiments with tasting the unfermented beer, setting up our own analysis lab, offering kegs of beer to UW researchers for use of their mass spectrometer, but in the end we ended up waiting for beers to ferment before we analyzed them. Because the cycle is so long we'd queue up all the data we could and some days we were tasting 18 batches of the same recipe "pale ale" with minor variations like the size of the hops container, what size mesh it had, etc. It took a while before we understood which variables really mattered and it often defied conventional brewing wisdom.
We also have had some really "interesting" batches, including a particularly remarkable scotch ale. Temperature control has always been great, but we had to develop a temperature probe that read accurately in the turbulent fluid stream and calibration procedures to make sure we got consistent readings no matter where the components come from. We had a machine that was reading off by a couple degrees and brewed a beer that smelled like beer and tasted like beer but had the texture of corn syrup. Jim was excited beyond words that we had done our extraction in such a narrow temperature that we got a very particular sugar profile with this texture. Nobody wanted to drink it, though.
How has the feedback been so far?
The feedback has been great.
First there were our friends, of course. They endured some hit or miss stuff in the very early days, but pretty quickly we were showing up at dinner parties with an unmarked growler and the first question was "Wow, what brewery is this from?" That's always a good feeling. In the last few months we've started inviting more professional brewers to taste our beers and try the system. It's interesting the different approach that different customers take when they are evaluating the machine. For the consumer audience there's a lot more about how the process works, is it really this easy, etc. For the pros, it's generally more of "Is the beer good or not?" and after that if we solved even a few of the problems we set out for it's a huge win for them. Fortunately the beer is great and by now we've had hundreds of people confirm this.
One thing that's interesting is which recipes really resonate with people. Paul Shipman from Red Hook talked about this in our video a little bit. We've got a wide range of "clone recipes" for commercial beers, some are published from the breweries, some are developed by homebrewers. We make it easy to take any recipe you find on the internet and adapt it to our system. But we've also had professional brewers come work on our machines for recipe development and those beers are typically more unique but still balanced in all the right ways. Those kegs always go quickly. Professional tools in the hands of a hobbyist can make something better than the hobbyist tools in the hands of a hobbyist, but professional tools in the hands of a professional is always a more promising combination.
Visit Kickstarter to learn more about the PicoBrew campaign!