Artisan Documentation

Visual scope for coffee roasters


Artisan meta philosophical guide

Artisan software can be understood as a component in a system of changing components. This system is graphically illustrated in the graph attached, which gives an idea of the place of artisan in roasting. The accompanying definitions describe the underlying meaning for each component and its relationship with other components. The graph tries to bridge separate perspectives into a coherent whole.

The study of coffee roasting is complex and divided. In science and engineering, it is assumed that the whole is the sum of each part. There is a tendency to divide and study things independently from each other. But roasting variables can create plural differences in other variables, producing complexity. One single cause can create several effects. For example, extending the amount of time at one temperature X, could create various different chemical compounds that affect the flavor.This complex nature would need a top level framework to try to understand it.

A system is a purposeful collection of interrelated components that work together to achieve an objective. The objective of artisan is to help make decisions that improve the taste of coffee. If a satisfactory experience of flavor is defined as coffee happiness. Then, the purpose of artisan would be to find and enable coffee happiness.

Solving a problem is transforming a given situation into a desired situation or goal. To solve a problem, in our case, to improve coffee flavor, a representation is usually generated, or a preexisting representation accessed. A representation need to include a description of the given situation, operators or actions for changing the situation, and a test to determine when the goal has been achieved. This is the where artisan fits in the system.

This chart uses the new Wheel Graph facility



This is the information collected. This is where artisan fits in the system. Artisan collects data from sensors for subsequent processing and analysis. Data can be divided in three types. Without data, it would be hard to make decisions.


This is the how of roasting. It describes roasting as a chemical and engineering process. If you do X, then Y happens. For example, at temperature X, time Y, acid concentration Z starts to go down. This would help make decisions that enable shaping the flavor.


This is the why of roasting. This is the cultural-human perspective and where decisions happen. Here we can set different goals and objectives like quality, taste, value, etc. Here, decisions are made based on a mix of cultural, rational, and emotional criteria. This is where divisions of goals occur. For example, a person could make the right decisions for her/him but maybe they would not be the right ones for someone else. This is the philosophical part of roasting.


In the process of roasting coffee, there are three categories of data, initial conditions (1), inputs (2), and outputs (3). Initial condition is data that describes the state of the roast before it starts. It can be weight, ambient temperature, roaster temperature, etc. For example, for two roast to taste the same, they would have to have identical initial conditions.

  • Inputs. This is data based on controlling changes. For example, changing the fan setting from a setting of 6 to a settings of 7 creates input data. Anything done to change or alter the process of roasting is recorded as input data. Since ET (Environment Temperature) is used as a controlling temperature for a desired BT (Bean Temperature) (after some time lag), ET is also input data.
  • Outputs. Since BT describes the final stage of the roast, BT is output data of the system. Cooling data would also be output data.
  • Behavior. This is how the beans respond to changes from an engineering and chemical perspective. The state of a roast could be ideally frozen at any particular time, as a sum of chemical and engineering variables. This knowledge is an enabler that bridges together taste (from an individual human point of view) and science. It helps to make decisions to achieve a goals.
  • Analysis. Here we observe the trends and their corresponding development in the flavor. We the data of previous roast to determine the steps needed to achieve our goals and create theories.
  • Model. The distribution of the BT temperature during a roast define a profile. A profile can be considered a roasting technique or a model, that imparts a flavor according to the beans used. Profiles can be used as roasting templates.
  • Decisions. Decisions can be cultural, rational, or emotional. This is the most human aspect of roasting. This is where we compare results with expectations. Decisions are based on degrees of freedom that in turn create choice. For example, if a roaster is not capable of controlling temperature, maybe the freedom of decisions would be restricted to time alone. Roasting decisions are determined by the level of personal satisfaction, flavor, choice, and knowledge.

2 Responses to “Philosophy”

  • Jens Oris says:


    Thanks for the great software, I’m using it know for almost 2 years and its getting better and better. Can I somewhere give further suggestions for developments in the future? Like a turning point timer for instance, this can maybe make the software even beter than it know is..

    With kind regards,


  • Marko Luther says:

    Thanks for being that faithful and for your suggestion Jens. We will work on this for the next release.