A Goal-Based Approach for Learning in Business Processes - Intentional Perspectives on Information Systems Engineering - page 237

Our learning approach seeks to identify the relevant contextual factors and group
process instances into context groups, such that for process instances of a specific
context group, similar process paths would imply similar outcomes.
3 Introducing the Running Example
As a running example, demonstrating the concepts introduced in the chapter, we
address a production process in a plastic bottle manufacturing firm, illustrated in
the BPMN model depicted in Fig. 1. The (hard) goal of the process is to reach a
state where customers' acceptance of delivery is achieved. Other states in which the
process might terminate (exception termination states) are states where delivery is
cancelled due to quality problems and states where the customer rejects delivery
(also due to quality problems).
Soft goals defined by the organization include increasing the percentage of deliv-
eries that meet their due dates, increasing machine utilization, reducing waste of
raw materials, increasing the quality of the manufactured products, and reducing the
overall production costs. These different soft goals could be prioritized and weighted
to form one composite soft goal. An alternative approach would be to analyze the
dependencies among soft goals and identify a dominant soft goal to be addressed
first. Table 1 presents the main causes for poor achievements of the defined soft
goals.
As seen in Table 1, the leading reasons for poor business results are quality
problems, machine failure and poor technical condition, and the set up operations.
Machine maintenance is not in the scope of this process (rather, it is part of its
context), and the set up operation is part of the process path. Based on this analy-
sis, we decided to focus on the soft goal of increasing the product quality, which
will affect all the other soft goals (including the costs, through reducing material
waste).
The contextual variables of the process include the initial case properties and
external (uncontrolled) events during the process. Initial case properties include
properties of the manufactured product and the customer, intended market of the
product (food, medical supplies, chemicals, cosmetics), the main raw material
(polyethylene at different density levels, polypropylene), the supplier of the raw
material (three possible ones), and the supplier of the pigments (two possible ones),
time since last maintenance operation of the machine, and weather (hot or dusty
days may affect the machines). There might also be specific requirements made by
the customer, such as requirements for the bottle to be resistant to high temperature
(in case the customer uses it for storing hot liquids) or to strong chemical solu-
tions. Events that occur during the process are mainly machine failures and quality
problems and it is often impossible to tell one from the other.
It should also be noted that contextual variables may affect soft goal thresholds,
e.g., higher quality is required if the intended market of the product is the medical
supplies.