Coffee Data Science
Coffee Roasting on the Roest
A first crack popping good time
I have been roasting coffee at home for 9 years on a Hottop, and over the past few years, I have been thinking about an upgrade. Luckily, my friends at Sweet Maria’s let me borrow their Roest to run some experiments, so I got to look at one of the most technologically infused coffee roasters on the market.
History
My Hottop has been fantastic, but it is the least data driven part of my coffee experience. I developed a roast profile based on taste for espresso over 8 years ago where I tried to balance too earthy coffee and medium roasts that taste burnt. I ended my roasts 1 minute after the first crack, and I experimented a little on this parameter as well. This development happened before I knew anything about roasting as well as years before I knew any professional coffee roasters.
I didn’t use a cupping routine to evaluate the roasts either. I would wait a day or two or maybe a week, and I would start pulling espresso on my lever machine. I did not let my roasts degas enough, and I was judging them based on how that espresso pulled. Additionally, I was roasting once a week, drinking that coffee for a week, and making a modification to the roasting profile.
With all of these deficiencies in process, after a year, I settled on a desired profile. The display on my Hottop promptly broke, and I could not make adjustments anymore. My Hottop was given to me used, and I was very thrifty at the time. I wasn’t going to change equipment unless my roasts weren’t good.
This roast profile stood the test of time as I improved my espresso shots and got to know more coffee people. Although, I longed for more data.
Roest
While attending the SCA Portland, I met Tom from Sweet Maria’s, and he said he had some roasters I could borrow if I wanted to try them out or collect some data. So I borrowed a Roest for a month.
In one month, I did 29 roasts. I got some beans from a lot of green coffee that had gone bad. So I wanted to iterate through a few roast profiles. I also roasted other coffee from Sweet Maria’s.
I fell in love with the machine quite quickly. It fit nicely on my counter, but it was also moveable for when I wasn’t roasting. I liked the whole process:
- Tweaking the roast profile
- Pour beans in slot above the roasting compartment
- Pulling the slot up to dump the beans into the roasting chamber
- Watching the bean color change
- Watching the roast plot
- Checking the color by pulling out some beans
- Pulling up the handle to dump the roast
- The square cooling tray with the light to view the finished roast.
In terms of interface, the web interface allowed for very quick profile modifications. This feature helped me complete back to back roasts will few issues.
The main challenge with roasting experiments is that you need a lot of green coffee and time. I have the former, not the latter with these tests. I suspect if done in earnest, it would still take me close to a year to make a major discovery. I’m still interested in doing such experiments because their complexity is very interesting.
Soon there will be a software update to the Roest where bean temperature and inlet temperature can be controlled at the same time. I’m excited for the possibilities of the Roest, and I look forward to the day when I can have one of my own.
If you like, follow me on Twitter, YouTube, and Instagram where I post videos of espresso shots on different machines and espresso related stuff. You can also find me on LinkedIn. You can also follow me on Medium and Subscribe.
