First Crack prediction


One of the Artisan features I miss is the prediction based on current temp and RoR of when FC is likely. At the moment I’m doing frantic mental arithmetic to figure out whether my current settings are going to over or under hit my target and obviously the earlier I can get a heads up, the easier it is to gradually make changes to stay on the glidepath.

If trend ahead plotting on the graph is too involved, a simple display box showing ETA to FC would be fine with me…

If there’s somewhere else to log enhancement requests, please point me in the right direction.


Hi Stuart
If you use the IBTS, then its 99% predictable as FC will start at the same temperature (for the same bean).
IBTS °C(F) / min is something we are working on, but even without it its very easy to see when you need to make changes before FC
The more we analyse the old type bean probe, the more we start seeing just how unpredictable it is.


Hi Jonas,

Maybe I was not clear in my explanation - I agree FC temperature is roughly predictable, what I was looking for is something like an ETA based on the current RoR etc. So if I am five minutes into the roast, the ETA for FC temperature might be 7 minutes and I’m aiming for 9 so I lower the power a little and watch the ETA change (much like a GPS navigation which would change your ETA to destination based on your speed). If I lower it too much say by six minutes I see my ETA has gone out to 10, I can increase a little to try and get back to the target.

At the moment I go by the readout on the R1, so say at 5 minutes in, my RoR is 15 and my current temp is 160. If my target is 9 minutes, then I’m estimating/predicting 4x15+160 = 220 which is too high, so I lower power and/or increase fan and then I recalculate in my head. It would be nice if the software on the laptop could keep doing that boring calculation at interval (15s?) and display it. I used the analogy of a pilot on an instrument glidepath coming in to land knowing he’s not too high nor too low for the runway based on his rate of descent. Something like that for arriving at FC temperature at a given time.

A display would be fine if amending the graph to project forward is too complex, but both would be nice :slight_smile:


Guess this is what all the projected lines were about in Artisan. Oops! Missed whole point! I was using my HotTop at the time and thermal response is not one of its strong suits.

Now that I understand what they were trying to tell me, I’m compelled to say “Me too!”



Late to the party in this topic, but yes prediction is very important information to have for those that are wanting to get ahead of what the beans may do in the future. The more hand/mind manual calculations that have to be done, the less time the user has to make changes before they are needed. Whether you roast totally manually or use live-software charting tools, the need for anticipating where a given bean is heading and making changes ahead of time remains. Intelligent software tools can make that work a lot easier, more effective and consistent.

Many of my roasts are over in just 11-12 minutes. When you consider the time-frame remaining after “drying”, the available time to make changes narrows. It can become a juggling/wishful act to make the right changes in advance of where they are needed during maillard. That is why Artisan has the predictive analysis included in the software as a “tool”. Intelligent prediction via the software can help reduce the “fog of ROR”, etc for the user. :sunglasses: