Roast level project

Roasting on the Bullet R1
#1

Hello all, was thinking about a starting a roast level project as I just got my Tonino colorimeter and figure with everyone on here it could be helpful/fun project to get data to approximately map out roast color levels of ground beans over different temperatures and times, and wanted to gauge if there were Bullet owners with colorimeters that would be interested in contributing to this project or if this has been done somewhere else.

I am a home roaster with almost a year of experience so my sample size is extremely small.

Please let me know if you have the equipment and would be interested in helping out and I can work on figuring out the execution approach if there is enough interest.

Best,

Kyle

4 Likes
#2

Interested in following your project.

#3

I think a few of us here use the Espresso Vision (aka Roast Vision). Since getting mine I’ve been trying to remember to take a reading of my roasts and log it in RW (sometimes I just write it down in my paper log and forget to transfer the data :expressionless: )

I guess what data are you trying to collect? % loss vs the color reading?

1 Like
#4

Thanks for reaching out. I am by far not the best at capturing my data. The primary goal of this project is to create a graph depicting roast levels for time and temperatures. For example knowing dropping a high elevation natural at 220C at 11 minutes would most likely result in a medium roast vs 220C at 7 would give a light roast.

I am initially thinking of working on capturing the time of the roast, temp (figure both probes if possible), color measurement, elevation, and washed vs natural, and batch size. I do like the idea of % loss being another field.

I know there are a lot of members that would be smarter than I am at different aspects of this project (and hopefully if it takes off people would be interested in providing guidance or take ownership of any aspect).

#5

Interesting premise to test total roast time v drop temp. I think the tricky part to this is each of us vary in how we roast. It’s a lot of variables and ambient temp is a factor too. Then you got to factor in the measurement device, your Tonino and my EV may call a “medium” slightly differently.

I think to make this manageable for you is do start with your own data to establish that pattern and your hypothesis :slight_smile: I’m not a data scientist but I’ve also been thinking about how to analyze my own data - the problem is right now there is no way to export all of my RT/RW meta-data into a csv file (at least) - the export we get now is all the raw data, one roast at a time (at least that’s how it is 6+ months ago).

#6

My thought is that with enough data some of the variables may not be as significant, and this would be more of an approximate than exact result. I have been thinking about the impact of charge temperature and heat application, but generally with the same batch size the result should be pretty similar and with enough samples the outliers would probably not have that much of an impact.

As for the measurement device, each of them do have alignments to Agtron values to compare apples to apples. I am also assuming they would be pretty similar in their measurements (I was looking into getting an EV but not sure they will ever be in stock again based on the email I got when asking).

The reason I’m enlisting the help of the community is I don’t roast much beyond the needs for my wife and I besides the occasional request of friends at this point, so it would take a long time to get meaningful results. Figuring if the few points were captured it would work well enough for purpose and not need the complete roast data.

#7

I wish there is an easy way for me to just share my roast data with you and you can run the analysis :wink: Unfortunately I don’t have much time for this even though I like the idea. I’ve only been once a month (but usually 2 to 3 batches each session) but I might be roasting a bit more now that I have some friends who wants to buy my beans… I guess they like it enough to want them :laughing:

3 Likes