Page 56 -

2 The Language: Rationale and Fundamentals 43
The next two patterns Event-based Task Trigger and Data-based Task Trigger also
provide mechanisms for synchronizing task commencement; however, they are
based on the occurrence of specified events such as the receipt of anticipated data
elements from the operating environment or from elsewhere in the process instance.
From both patterns, either a delayed task instance can be allowed to commence or a
completely new task instance can be triggered.
Finally, the Data-based Routing pattern denotes the situation where the value of
one or more data elements can affect the flow of control in a process. It aggregates
two control-flow patterns – Exclusive Choice and Multichoice – where data elements
are used as the selection mechanism for determining whether the thread of control
should be routed down a given branch in the process.
While the data patterns focus on identifying the various ways in which data
impacts on and is distributed by a business process, the resource perspective delin-
eates the ways in which resources, both human and non-human, are involved in the
conduct of work associated with business processes.

2.2.3 Resource Patterns

Resource patterns characterize the way in which work is distributed to the resources
associated with a process and managed through to completion. There are seven
distinct groups of patterns that describe the various aspects associated with the
involvement of resources in automated business processes. These are listed below.
Creation patterns describe design time work distribution directives that are
nominated for tasks in a process model.
Push patterns describe situations where the system proactively distributes work
items to resources.
Pull patterns relate to situations where individual resources take the initiative in
identifying and committing to undertake work items.
Detour patterns describe various ways in which the distribution and life-cycle of
work items can deviate from the directives specified for them at design time.
Auto-start patterns identify alternate ways in which work items can be automat-
ically started.
Visibility patterns indicate the extent to which resources can observe pending
and executing work items.
Multiple-resource patterns identify situations where the correspondence bet-
ween work items and resources is not one-to-one.
One of the major determinants of the way in which individual resource patterns
apply is related to the state of a given task instance that a resource may encounter
at runtime. Figure 2.12 illustrates the life-cycle for a task instance or work item
as they are often known in a workflow context. The diagram indicates the various
states through which a work item passes during execution and the arcs indicate the