How hard is roasting? High end programmable home roasting machines?

#1: Post by **dsc106** » December 18th, 2021, 2:33 am

Haven't thought about roasting my own green beans much - sounds time consuming and like a whole additional can of worms, no thanks! I'm surrounded by awesome local micro roasters.

But I'm also aware that coffee prices are going to spike, so contemplating if it's worth it to have a roaster in my garage and just roast myself.

How difficult is it to get results that match great micro roasters at home?

Are there high end home machines - like a Decent Espresso style machine - for roasting? Where I can use software to match other people's profiles to easily create ideal roasts for different varietals, and then tweak myself to taste? And to be able to do it pretty hands off?

I've seen the sandbox smart R1 and it looks interesting, if not a little fresh off the boat. But can home roasting green beans compare to top tier micro roasters?

#2: Post by **_Ryan_** » December 18th, 2021, 4:11 am

I've been wondering the same thing, the https://kaffelogic.com/home/nano7/ is meant to be pretty good and appears to do what you want, and the https://ikawahome.com/ was recently released albeit with advanced settings locked behind a subscription.

I worked out that a Kaffelogic would "pay" for itself within 12 months (be cost neutral by month 12), and I'd be well ahead in year two using it for every day drinking, even still buying specialty beans when small batches take my fancy.

The small batch size does put me off, as I drink 20g shots and 17-20g pour-overs.
I assume six months or so to learn the basics, or if buying the machine locally from someone who also sells greens, I'd be expecting some profiles(there is a local vendor who does this I believe).

Watching this thread with interest.

#3: Post by **luca** » December 18th, 2021, 5:02 am

But can home roasting green beans compare to top tier micro roasters?

Totally possible!

How difficult is it to get results that match great micro roasters at home?

I guess the devil in the detail here is what your definition of a "great" micro roaster is, but I suppose I can speak to my experience. I recently home roasted some coffee that I was very happy with, and which I feel stands up to the quality of some of my favourite roasters. Getting the roasting right only took me going through a popper, heat gun, gene cafe, hottop, quest, kaffelogic, modified quest with pid controlled hot air feed and artisan hook up, landing on a ROEST, about a thousand logged roasts, ten years of experiments and help from about twenty engineer and coffee industry friends (who I won't name to spare them having people asking them for unpaid consultancy help), not counting the unlogged roasts, decade or so that I worked in coffee before then and the occasional sample roasting that I did. Getting good green coffee this year only took me rejecting about 50-100 samples (not counting the 20 years of experience tasting coffee and doing the Q grader course several times that it took for me to learn how to do this).

I worked out that a Kaffelogic would "pay" for itself within 12 months (be cost neutral by month 12), and I'd be well ahead in year two using it for every day drinking, even still buying specialty beans when small batches take my fancy.

At, say, $30/kg for home roasted coffee as all-in roasted costs instead of $80/kg for roasted coffee, I'd guess this is saving $50/kg. The coffee I experimented with probably had an average price of $20/kg, so if I use, say, 100g of home roast a week (I'll still buy commercial coffee), the savings will pay for all the green coffee that I used to get to this point in only about 7.7 years and then my savings will start to "pay" for my equipment! I haven't done the maths, but I'm in my 30s, so I probably have sufficient life expectancy to get to cost neutral.

Also, I should confess that I have only been able to roast Kenyan coffees to a level I am very happy with; everything else needs work. (I thought I should add that disclaimer in case anyone thought I was boasting too hard about how easily this came to me.)

Have fun!

#4: Post by **Almico** » December 18th, 2021, 8:42 am

That's like asking how hard is it to play chess. Pretty easy to learn the moves... Many years to learn how to do it well.

The general consensus among home roasters would agree that roasting to save money is not a great idea. The only good reason to roast coffee is if you learn to love roasting coffee.

The Roest seems to be the machine you are looking for to automate your roasts. It's about $5000USD. Should pay for itself in 10-15 years.

#5: Post by **Randy G.** » December 18th, 2021, 9:19 am

dsc106 wrote:Haven't thought about roasting my own green beans much - sounds time consuming ..

Mill City has some very good YouTube video lessons. In one they say, "If the boss asks me what did I learn roasting today, and I can't answer, I should be fired."

Over the last 22 years I have been roasting my own coffee for about 20 of them. Even worked for a roaster manufacturer. Been roasting on a new-to-me gas roaster for the last 30 batches. I was just gettin a feel for the machine and two batches ago upped the mass from 500 to 700grams. It was very uneven. Learned that I needed to increase the drum speed.

It is a form of cooking that is different from just about any other. I would give you the same answer when someone sees me walking my German Shepard and asks, "I have always wanted a German Shepherd. Do you thik I shouild get one?" I always answer, no. If you have to ask it is time do do more research. And you really have to WANT to roast.

#6: Post by **MNate** » December 18th, 2021, 11:03 am

I think you can reasonably save $5/pound roasting on your own and getting results you like. The pros roast better but you can roast to the color you like so it kind of evens out. Pros have better selection of beans but then when it actually gets to you you have less selection because you only get to choose what they've curated. Buying your own greens you have more selection. Too much, really.

I think my Fresh Roast broke even in the 2 years I used it, but it was a lot of work dragging out and took constant attention as it roasted. Think half hour per pound.

The Ikawa Home that some of us have recently bought takes almost no time as it does just what you said- roasts a profile on its own as you do other things. And $1k is a good value for that, I believe. The results are good. But at $5/pound savings it will take quite awhile to break even. Not worth it for cost.

Where all this might be worth it is I do think I've learned a lot more about coffee while doing it. Seems like a reasonable step in a hobby. And always the comparison to other things: I just spent $800 last night on Nordic ski stuff for my daughter... and I don't even know if she'll like skate skiing!

#7: Post by **pcofftenyo** » December 18th, 2021, 11:22 am

Another pretty high end programmable machine is the Bullet. It allows for sharing of roast profiles and playback of recent roasts which is basically automated roasting.

There is a thread hear about it: Bullet roast and learn.

Broadly speaking though is that yes, it is possible to match or even sometimes beat professional micro-roaster results as not all of them are good. My local micro-roaster is LOUSY!

It does take some time and some intention. High quality beans aren't hard to come by at all in the US and there are threads here for it too.

It has saved me a LOT of money since I've been roasting for over 7 years. Specialty that I purchase runs $1.50 - $2.00/oz (not super high end but good) and I buy equivalent green for $0.38-$0.60 pretty regularly. Plus we drink 1.25 to 1.5 pounds/week.

Luca paints where we can end up. That's clearly an option but not a requirement. In fact, I may need to hit him up for some consultancy!

#8: Post by **sfdamon** » December 18th, 2021, 12:54 pm

MNate wrote:I think you can reasonably save $5/pound roasting on your own and getting results you like.
...
I think my Fresh Roast broke even in the 2 years I used it, but it was a lot of work dragging out and took constant attention as it roasted. Think half hour per pound.

Same for me. My first roaster was a FreshRoast, I used it every weekend for about two years. I felt it paid for itself pretty quickly even though I was brand new to roasting. It's an easy machine and I recommend it to anyone curious about roasting.

Eventually I got tired of running so many small batches and I upgraded to a Bullet. Bullet probably pays for itself after about 500 pounds.

December 18th, 2021, 12:54 pm

luca's response is what I had envisioned, and what I was afraid of.

The thing is, I *don't* want to roast. In fact, I want to *not* roast. But I'll tell you what I do want, I do want to pay 40% of the price for coffee that is as good or better than the best, with almost zero effort. I want to put green beans into a smart digital machine and hit the "make it awesome" button. I want to make roasted coffees that would make world champion baristas swoon by accident. I want to ride off into the sunset (on a horse I didn't have to learn how to ride), sipping my macchiato, laughing all the way to the bank.

I want the ROEST Machine, v2.0. Why not v1.0, what's wrong with v1.0? I don't know, but something, something that someone on a forum somewhere knows. And I want that thing fixed, and I want new features added that would surprise even Elon Musk & Scott Rao alike, features that are over my pay grade that I don't even understand. But I want them, and I want the cost of the machine to be under $1000.

I want to go onto a digital ordering platform that has a reddit like upvote/downvote system where the world's best green beans right now have already been sampled and approved by people better at this than me, so I can circumvent the years of effort the best roasters go through to curate relationships with roasters - all by lifting my index finger. Nay, my pinky finger. And then I want to click buy and have it delivered to my front door, where I then walk a minimal distance to my all electric device, put into the machine without even reading the manual, and having people ask me how I did it.

When I roast a bean and I don't quite love the results, I want to hit the "make it juicier" or the "more chocolate" or the "bump dat sweetness" button, in which it will automatically reference the metadata and collective results of geniuses of the internet community, alongside the complex science of roasting, and know exactly how to adjust my Ethiopian bean to achieve the results I am after. I want a library of internet profiles, also with a reddit upvote/downvote like system and comments, where I can select different top rated roast profiles for my specific bean, curated from people who really knew what they were doing.

I want to send my kids to college on the money I saved, even though I only drink 20g/day and am the only coffee drinker in the house.

Now, I know what you are thinking. "Oh, you're being facetious, that's kind of cute." WRONG. I am being serious, DEAD serious.

Ok, maybe not dead serious. Only about 90% serious. And for all the jest, I can't imagine we're potentially more than a few years away from something like this. Last year I got a Masterbuilt Gravity Smoker, with about 10 hours of reading, a few rounds of practice, I was smoking meat as good as people who had done this for 20 years with near minimal effort, because the PID does all the hard work, and the gravity hopper does all the dirty work, and the internet has all the tried and tested answers to questions that prior generations worked so hard on. A few years ago I got a Sous Vide, and have been spoiled by the ability to make precision perfect meat. And then I realized you could combine Sous Vide & Smoking, with a little help from our friend J Kenji Lopez, and can make world class brisket by properly combining the two with minimal effort.

And for all the talk of Espresso as an art, outside of latte art, I think we're getting wise to the fact that it's not voodoo - it's math and science, with a hint of artistic customization at the end. Think of it, in 10 years an advanced Decent machine will be a TRUE super automatic, reading puck resistance, flow rate, TDS, EY, temperature, etc. 1000s of man hours of work will distil into several profiles for different bean types with automatic smart sensors to adjust for your bean age, storage conditions, CO2 levels, ambient temp, humidity, elevation, and personal preferences. Metadata on a bag will link with elevation and region and process, along with large communities of people who have upvoted the best profiles, and software which will be intuitive enough to make even complex customizations easy, even though what is happening behind the scenes is mind boggling.

Ok, where I am going with this? I DONT KNOW. But I was hoping that in 2022-2025 timeframe we may reach the point where a machine like the ROEST could be had for under $1k, and the power of digital communities and algorithms would enable stellar results with perhaps a 10 hour reading/experimenting investment to get me 90%-95% of the way to "as good as many top roasters", with another year of practice to be THE BEST IN THE WORLD.

As far as who my local micro roasters are, well, competition is steep - I'm in the PNW with names that would make global roasters blush, and I'd consider any roaster that can ship to my front door within 2 days of roast "local", so, there's that.

Thanks for humoring my tongue in cheek post, internet is hard, read it all with a wink and a smile. Sounds like I'm asking for too much today, but in a few years I wonder if this post won't somehow suddenly seem (semi)-reasonable??

#10: Post by **Peppersass** » December 18th, 2021, 2:16 pm

dsc106 wrote:Thanks for humoring my tongue in cheek post, internet is hard, read it all with a wink and a smile. Sounds like I'm asking for too much today, but in a few years I wonder if this post won't somehow suddenly seem (semi)-reasonable??

I doubt it. Roasting coffee resembles cooking to some extent, but automating the process is vastly more complicated. This is partly due to differences between different cultivars, different sources of the same cultivar, and even different lots from the same source. And those differences change from year to year. And there are also changes as the greens age. Ambient temperature and humidity make a difference as well. If that's not enough variables, different brew methods often require different roasting parameters. And the real killer complication is that human taste preferences for coffee vary greatly. I think automated roasting like you describe would require thousands or millions of parameter combinations that would require much programming to get right for any given green, brew method and person. Hardly push-button stuff.

The Ikawa 100g Home attempts to automate the process, but it falls short of what you suggest. They supply you with greens and a profile that supposedly produces a great roast. But what's the difference between that and simply buying roasted coffee? In both cases, you're relying on what the vendor thinks is a great roast. Ikawa does give you a few parameters you can alter to achieve a desired color and/or development time, but then you're back to experimentation, trial-and-error, etc. Hardly a push-button experience. The advanced profile editing offered by subscription only takes you down a deeper rabbit hole. I'm not knocking the Ikawa 100g Home. I think it's great, but it it's not automated to the degree you desire.

It took me almost no time to learn how to Sous Vide, but that's because it's a really simple technique. The only variables are temperature and time, and there are many published guidelines for that. In fact, Sous Vide tends to eliminate the "agricultural differences" between products. I have yet to find a steak that doesn't turn out perfectly when I Sous Vide it at 133 degrees for one hour.

Smoking meat might be *slightly* closer to roasting coffee. It can be difficult to do, but it's nowhere near as complicated. It's taken me about 15 years to learn how to smoke meat that makes people swoon, but even my earliest attempts turned out pretty well, even though back then I used a small Weber Smokey Mountain (I've since graduated to a ceramic Kamodo Joe cooker, which works way better in sub-freezing temperatures.) Basically, you just have to keep it low and slow, and maintain a constant temperature. Modern PID-controlled fans make this trivial to do. But there aren't all that many variables. The biggest issue is figuring out how long it'll take to smoke a given piece of meat (say, a 17 lb brisket, the most difficult cut to get right.), and the variation in shape and fat content between two hunks of meat, smoker characteristics, ambient temperature and wind probably defy any attempts to figure that out by computer to anything closer than an hour or two (or more.) Luckily, smoked meat can be held for hours in a cooler before serving, but meal timing is still a great source of anxiety for me.

Bottom line, I very seriously doubt that coffee roasting will ever be automated to the degree you suggest, even with AI far more advanced than we have now (and by then AIs will have replaced the human race anyway, and they won't be drinking coffee.)

And to pour salt in the wounds, it's unlikely that home roasting will ever save you much money, if any. Aside from the cost of a good roaster and your time, these days the very best greens in small quantities cost almost as much as roasted beans -- $15-$20 per pound (though a pound is 16 ounces instead if 12 oz!)