Patterns and Products

1
Patterns and Products
Wil van der Aalst

• Eindhoven University of Technology
• Faculty of Technology Management
• Department of Information and Technology
• P.O. Box 513
• 5600 MB Eindhoven
• The Netherlands
• w.m.p.v.d.aalst_at_tm.tue.nl

See www.workflowpatterns.com!
2
Selection of workflow management systemsAn
approach based on patterns
3
Workflow patterns initiative www.workflowpatterns
.com

• Started in 1996, joint work TU/e and QUT (1999).
• Different types of patterns
• Control-flow patterns
• Data patterns
• Resource patterns
• Some of the people involvedArthur ter Hofstede
  (QUT), Marlon Dumas (QUT), Nick Russel (QUT),
  Petia Wohed (DSV), Bartek Kiepuszewski (QUT),
  Alistair Barros (SAP), Oscar Ommert (EUT), Ton
  Pijpers (ATOS), Nataliya Muylar (EUT), Maja Pesic
  (EUT), Alexander Norta (EUT), Eric Verbeek, et
  al.

4
20 Control flow patterns

• FocusThe routing of work from one activity to
  another, i.e., a partial order of activities.

5
Classes of control-flow patterns

• Basic control flow patterns the five most basic
  patterns
• Advanced branching and synchronization patterns
  more complicated splits and joins
• Structural patterns absence of syntactical
  requirements that limit the modeler
• Patterns involving multiple instances e.g.,
  order lines in an order
• State-based patterns states are more than
  queues in front of activities
• Cancellation Patterns removing work-items and
  cases

6
Basic control flow patterns Pattern 2 Parallel
Split
7
Basic control flow patterns Pattern 3
Synchronization
simple right??
8
Basic control flow patterns Pattern 4 Exclusive
choice
XOR-split
9
Basic control flow patterns Pattern 5 Simple
merge
simple right??
10
It's not that simple!

• There are many ways to split!
• There are many ways to join!

11
Exclusive choice versus Deferred choice
Not that simple 1 The moment of choice matters!
12
Example Deferred Choice (Pattern 16)
process_form
send_form
c1
c5
archive
time-out
c3
evaluate
start
register
ready
c7
c2
c6
c4
check_proc
process_complaint
13
Not that simple 2 There are many ways to join!
A
C
join
B

• COSA (Ley) Places have capacity 1.
• Oracle BPEL/FLOWer True and false tokens.
• Staffware (TIBCO) Race conditions.
• SAP Business Workflow (SAP) split-join block
  with a minimum or an end condition.
• Etc.

14
Staffware Unintended race conditions
15
More patterns ...

• Advanced Branching and Synchronization Patterns
• Pattern 6 (Multi-choice)
• Pattern 7 (Synchronizing Merge)
• Pattern 8 (Multi-merge)
• Pattern 9 (Discriminator)

• Basic Control Flow Patterns
• Pattern 1 (Sequence)
• Pattern 2 (Parallel Split)
• Pattern 3 (Synchronization)
• Pattern 4 (Exclusive Choice)
• Pattern 5 (Simple Merge)

• Structural Patterns
• Pattern 10 (Arbitrary Cycles)
• Pattern 11 (Implicit Termination)

• State-based Patterns
• Pattern 16 (Deferred Choice)
• Pattern 17 (Interleaved Parallel Routing)
• Pattern 18 (Milestone)

• Patterns involving Multiple Instances
• Pattern 12 (Multiple Instances Without
  Synchronization)
• Pattern 13 (Multiple Instances With a Priori
  Design Time Knowledge)
• Pattern 14 (Multiple Instances With a Priori
  Runtime Knowledge)
• Pattern 15 (Multiple Instances Without a Priori
  Runtime Knowledge)

• Cancellation Patterns
• Pattern 19 (Cancel Activity)
• Pattern 20 (Cancel Case)

16
Evaluation
17
39 Data patterns

• FocusData patterns aim to capture the various
  ways in which data is represented and utilized in
  workflows.

18
Classes of data patterns

• Data visibility relating to the extent and
  manner in which data elements can be viewed by
  various components of a workflow process.
• Data interaction focusing on the manner in
  which data is communicated between active
  elements within a workflow.
• Data transfer which consider the means by which
  the actual transfer of data elements occurs
  between workflow components and describe the
  various mechanisms by which data elements can be
  passed across the interface of a workflow
  component.
• Data-based routing which characterize the
  manner in which data elements can influence the
  operation of other aspects of the workflow,
  particularly the control flow perspective.

19
Data visibility Pattern 3 Scope data
20
Data interaction Patterns 14-17
21
Resource patterns
22
43 Resource patterns

• FocusThe gap between the creation of a
  work-item (activity that needs to be executed)
  and the actual execution by some resource.

23
Classes of resource patterns

• Creation patterns design-time work allocation
  directives
• Push patterns workflow system proactively
  distributes work items
• Pull patterns resources proactively identify
  and commit to work items
• Detour patterns re-routing of work items
• Auto-start patterns automated commencement
• Visibility patterns observability of workflow
  activities
• Multiple resource patterns work allocation
  involving multiple participants or resources

24
Push patterns Pattern 17 Shortest Queue
push workitem to shortest queue
25
Pull patterns Pattern 23 Resource Initiated
Exec.
pull workitem to start working
26
Auto-start patterns Pattern 38 Piled execution
the completion of one activity starts another
cf. Pattern 39 Chained execution
27
FLASH Animations of Patterns
28
Evaluation of 5 concrete WFM systems
29

• Four types of "workflow-like" systems
• Information systems with hard-coded workflows
  (process organization specific).
• Custom-made information systems with generic
  workflow support (organization specific).
• Generic software with embedded workflow
  functionality (e.g., the workflow components of
  ERP, CRM, PDM, etc. systems).
• Generic software focusing on workflow
  functionality

SAP Business Workflow/Webflow (SAP AG)
FLOWer (Pallas Athena)
Oracle BPEL (Oracle)
Staffware (TIBCO)
COSA (COSA GmbH)
30
SAP Business Workflow/Webflow (SAP AG)
Staffware (TIBCO)
Oracle BPEL (Oracle)
COSA (COSA GmbH)
FLOWer (Pallas Athena)
31
Staffware (TIBCO)
32
Staffware Control flow perspective
33
AND-split
OR-join
34
Example
35
Staffware Control flow evaluation

• Classical problems
• no states
• only internal choices
• Evaluation based on
• Workflow Patterns (2002)
• new features (prediction/fragments) do not
  influence evaluation

36
Staffware Data perspective

• Ability to define forms
• Variables of different types
• Evaluated in Workflow data patterns 2005.
• Weaker than most other systems.

37
Staffware Resource perspective

• Concept of work queues
• Evaluated in Workflow resource patterns 2005.
• Weaker than most other systems.

38
Staffware Conclusion

• Negative
• Limited functionality in all perspectives.
• Little flexibility.
• Future within Tibco is less clear.

• Positive
• Widely used
• Easy to use
• High performance
• Interesting new concepts, e.g., prediction and
  fragments

39
COSA (COSA GmbH)
40
COSA Control flow perspective

• Based on Petri nets
• Lots of functionality

41
COSA Control flow evaluation

• Benefits from Petri net basis.
• Support for advanced synchronization and multiple
  instances patterns could be better.
• Evaluation based on
• Workflow Patterns (2002)

42
COSA Data perspective

• Variables at different levels.
• Good support through tool agents.
• Evaluated in Workflow data patterns 2005.
• Weaker than most other systems.

43
COSA Resource perspective

• One of the most powerful tools on the market.
• Multiple dimensions and rules.
• Evaluated in Workflow resource patterns 2005.

44
(No Transcript)
45
COSA Conclusion

• Negative
• No integrated forms environment
• Little operational flexibility
• Future of COSA is less clear

• Positive
• Powerful language
• Unique features in each of the perspectives

46
Oracle BPEL (Oracle)
47
Oracle Control flow perspective

• Based on BPEL

48
BPEL(4WS) Business Process Execution Language
for Web Services

• Compromise between IBM and Microsoft.
• Merges WSFL and XLANG.
• Two styles of working graph based and
  structured.
• Expressive but complex language.
• Support by many vendors IBM Websphere, Oracle
  BPEL, and at least 16 more products.
• In April 2003, BEA Systems, IBM, Microsoft, SAP
  AG and Siebel Systems submitted BPEL4WS 1.1 to
  OASIS.
• Version 2.0 will have subtle but significant
  differences.
• Two flavors executable and abstract
  (non-executable)

49
Constructs

• Primitive activities
• invoke, invoking an operation on some web
  service
• receive, waiting for a message from an external
  source
• reply, replying to an external source
• wait, waiting for some time
• assign, copying data from one place to another
• throw, indicating errors in the execution
• terminate, terminating the entire service
  instance and
• empty, doing nothing.

• Structured activities
• sequence, for defining an execution order
• switch, for conditional routing
• while, for looping
• pick, for race conditions based on timing or
  external triggers
• flow, for parallel routing and
• scope, for grouping activities to be treated by
  the same fault-handler.
• Activities can be nested.
• Can be connect though links.

50
Oracle Control flow evaluation

• Support similar to BPEL (2 exceptions).
• Evaluation based on
• Pattern-based Evaluation of Oracle-BPEL
  (v.10.1.2), N.A. Mulyar, 2005.
• few ratings have been changed to make a fair
  comparison possible

51
Oracle Data perspective

• Similar to BPEL, e.g. scope (create and scope
  data), assign (copy and modify), invoke (exchange
  with external entities), etc.
• XML based.
• Powerful but complicated.

See Pattern-based Evaluation of Oracle-BPEL
(v.10.1.2), N.A. Mulyar, 2005.
52
Oracle Resource perspective

• Not part of BPEL (cf. BPEL4people).
• Oracle specific task implementations.
• Close to programming.

See Pattern-based Evaluation of Oracle-BPEL
(v.10.1.2), N.A. Mulyar, 2005.
53
Oracle Conclusion

• Negative
• Based on BPEL (close to programming and a bit
  immature)

• Positive
• Based on BPEL (emerging industry standard)
• Powerful language

54
SAP Business Workflow/Webflow (SAP AG)
55
SAP Control flow perspective

• Two views native view and EPC view.
• Block structured.
• e.g. sequence

56
Routing elements
57
parallel routing
three types of choices
58
join can hove a condition and/or lower bound
two types of loops
59
SAP Control flow evaluation

• Classical language with classical capabilities.
• See Pattern-based Evaluation of SAP Business
  Workflow, 2005.

60
SAP Data perspective

• Data is stored in containers holding values or
  references to (SAP) objects (e.g., documents,
  masterdata, etc.).
• Workflow containers and task containers are
  linked through bindings.

61
SAP Resource perspective

• Organizational units have positions that may or
  may not be occupied.
• Positions may be associated to multiple jobs
  (kind of role).
• You can assign a task to
• an organizational unit, if it is to apply to all
  subordinate positions
• a job, if it is to apply to all positions
  described by the job
• a position, if it is to apply to those persons
  (employees) or users who hold the position
• a person (employee), if it is to apply to this
  person

62
Logging in SAP Workflow
63
SAP Conclusion

• Negative
• Integration in SAP environment (for those that do
  not have SAP).
• Limited functionality in all perspectives.
• Little flexibility.

• Positive
• Integration in SAP environment (for those that
  have SAP)
• free
• support
• solid future (although technology may change
  towards BPEL)

64
FLOWer (Pallas Athena)
65
FLOWer Control flow perspective

• Block structured (similar to BPEL) but with
  excellent support for multiple instances.
• Can be generated from Protos.
• Case handling principle.

66
Partly block structured choices, loops, etc.
require an additional level
67
FLOWer Control flow evaluation

• Powerful language.
• Many /- because of case handling concept.
• Evaluation based on
• Workflow Patterns (2002)
• more recent releases also support synchronizing
  merge.

68
FLOWer Data perspective

• Case handling concept data driven.
• Forms play inportant role.
• Multiple instance data.

Evaluated in Workflow data patterns 2005.
69
FLOWer Resource perspective

• Three roles for step in the process (execute,
  redo, skip).
• Hierarchies can be defined independent of
  process.
• Various was to view/distribute work.

• Evaluated in Workflow resource patterns 2005.

70
Forms
71
(No Transcript)
72
FLOWer Conclusion

• Negative
• Complicated organizational model and more work
  needed to restrict people (if needed).

• Positive
• Highly expressive language
• More flexible
• Powerful forms environment
• Strong presence in The Netherlands

73
Other aspects important for evaluation

• Price
• Reliability/stability vendor
• Product support (training, etc.)
• Implementation support
• Performance (load and response times)
• Application integration
• Technical constraints (e.g., databases, etc.)
• ...