This page is now obsolete: it was intended to help you install and use IBM's
WebSphere Business Integration Workbench, a part of the
WebSphere Business Integration Modeler
(itself part of the IBM WebSphere® MQ
Workflow, formerly named "MQSeries® Workflow") that used to be freely available
for 30-day trials.
Note that the WBI allows you to represent some workflow graphs but not genuine Petri nets. An example is given at the end of this file. The graphs here have no explicit token (and hence no multiple tokens).
Table of Contents
The downloading and installation
is straightforward if you use the default installation options
(ignore the references to "the CD", just click "Ok" whenever asked to).
However then, you need to be logged as "administrator" on your PC.
The downloading creates the directory "C:\downloadDirector". Then,
if you double-click on "IBM-WBI-Workbench-V424-Trial-Edition.exe" in this
directory, the directory "C:\BPM" is created and the workbench installed in it
(if you click "ok" whenever asked to).
Then, you can run the workbench by double-cliking on "BPM Workbench.exe" in
"C:\BPM" (if you do not see the extension ".exe", click on the first file named
"BPM Workbench"; the second file with that core name just stores an icon).
When opening sample process descriptions, you may be asked for some
currency unit conversion details: select "Australian dollar" in the right part
of the provided menu, then "US dollar" in the left part, then the "<-"
arrow and finally click on "Ok".
The documentations are in "C:\BPM\Docs".
Getting Started (86p).
Chapter 1: installation of the workbench and then presentation of the main menus.
Chapter 2: mini-tutorial (20 min), an abstract of the "Tutorial".
Except to get a feeling of the interface by following this mini-tutorial a bit,
I do not advise to "get started" with that guide since it lists a lot of things
but gives very few explanations and hence does not permit to "understand"
the tool. Thus, there is no need to print this guide.
Tutorial (134p). This guide provides a hands-on practice session which supposedly takes four to six hours to complete but you may actually take this time just to read it. If you have got a feeling of the interface (as suggested above), I'd suggest to print this tutorial and read it once or twice before testing it: one reading to get some idea of all the proposed functions and their interaction, and then armed with this high-level view, a second reading to really understand the functions and their interactions. Read the annexes as soon as they are refered to in the text (you may also read them all at once). Below, I have gathered the main process descriptions in this tutorial in order to provide a much needed synthesis (I also illustrate a textual/linear format to permit the description of processes within e-mails). If you want to limit printing, print only the annexes and the images below. In this tutorial, the chapter 2 is about creating object types while the chapter 3 is about creating and connecting objects (entities, phis, organization units, roles, tasks, ... ; see the glossary at the end of the tutorial if you forget their meanings).
The other guides provide additional explanations and present additional
functions.
The Analysis Guide (90p) is helpful
to better understand and used simulation and weighted average.
The Modeling Guide (272p) repeats and
extends what is in the tutorial (analysis excepted).
The User Guide (406p) seems to present
a bit everything.
Other proposed guides:
the Modeling Business Measures
(78p),
the Integration Guide (314p; about
integration of the modelling with actual workflow systems, including the MQ worflow
system), the Reporting Guide (142p),
the Release Notes (18p),
the UML Modeler User Guide (210p)
and the XForm Designer User Guide (64p).
The "Sales Order Fulfillment As-Is" Process (p 69 of the Tutorial):
In the linear format that I propose for describing processes within e-mails:
[Receive Order As-Is /Order Processing]
{ ->(Work Order /Paper Form)->[Packaging As-Is /Packaging]
{ ->(Work Order /Paper Form)->[Shipping As-Is *s /Shipping],
->(Packaging Material /Packaging parts)->[*s]
}
->(Work Order C1 /Paper Form)->[Product Assembly As-Is /Assembly]
{ ->(Ordered Product /Product)->[*s],
->(Work Order c1 /Paper Form)->[*s]
}
};
[*s]->(Shipment Confirmation /Paper Form)->[Generating Invoice As-Is
/Accounts Receivable].
//Note 1: the part after a '/', i.e. the media, is optional.
//Note 2: an alternative format for the last statement is
// [Generating Invoice As-Is]<-(Shipment Confirmation)<-[*s].
//Note 3: you can write everything in lowercase if you prefer.
The "Receive Order As-Is" Subprocess (pp 35 and 38 of the
Tutorial):
In the proposed linear format (without media specifications):
(Customer| Phone Order)->[Enter Order Information]
->(Sales Order)->[Check Credit History]->< Customer on File? >
{ yes 70%->(Credit Report)->[Approve Customer Credit *a],
no 30% ->(Customer Credit Information)->[Run Standard Credit Check]
->(Faxed Credit Report)->[Forward to Account Receivable]
->(Faxed Credit Report)->[*a]
};
[*a]->(Notification of Credit Approval)->[Approve Customer Credit]
->< Order Approved? >
{ yes 75%->(Sales Order)->[Create Work Order *a] { ->(Work Order),
->(Work Order Ct) },
no 25% ->(Sales Order)->[Reject Customer Order]->(Stop)
}.
//Note: the above is not a genuine Petri net since within a Petri net
// a decision (explicit OR-split) is a "task" (transition) and a task
// must be followed by a "place" (condition); hence for a Petri net,
// "Customer on File?" should normally be square bracketed and a place should
// be inserted between [Check Credit History] and [Customer on File?];
// alternatively, these two tasks may also be merged.