th=10 nowrap>Yahoo! is not affiliated with the authors of this page or responsible for its content.
Reducing OEM Development Costs and Enabling Embedded Design Efficiencies Using the Unified Modeling Language (UML 2.0)
Reducing OEM Development Costs
and Enabling Embedded Design
Efficiencies Using the Unified
Modeling Language (UML 2.0)

Getting to Market Faster with Lower Cost







Jerome L. Krasner, Ph.D.
February 2004


EMBEDDED MARKET FORECASTERS

American Technology International, Inc.

© 2004 Embedded Market Forecasters

Copyright 2004 by Embedded Market Forecasters, a division of American Technology International, Inc,
1257 Worcester Road #500, Framingham, MA 01701. All rights reserved. No part of this document covered by
copyright hereon may be reproduced or copied without expressed permission. Every effort has been made to
provide accurate data. To the best of the editors knowledge, data is reliable and complete, but no warranty is
made for this.
Reducing Embedded Development with UML2.0
2 of 26

© 2004 Embedded Market Forecasters

Table of Contents
Reducing OEM Development Costs and Enabling Embedded Design
Efficiencies Using the Unified Modeling Language (UML 2.0)
.........................1
Overview ...............................................................................................................................4
Looking at Embedded Design Processes..............................................................................6
UML 2.0 to the Rescue..........................................................................................................9
Static Diagrams .................................................................................................................9
Dynamic Diagrams ..........................................................................................................10
Behavioral Diagrams .......................................................................................................11
Going Beyond the Standard ................................................................................................11
UML 2.0 Tools of Importance ..............................................................................................13
How Vendors Differentiate Their Tools While Supporting UML 2.0 .................................14

Model-Driven Development Productivity Enhancers ...........................................................16
Production Code Generation ...........................................................................................16

Model/Code Associativity (Roundtrip engineering) ..........................................................17

Reverse Engineering .......................................................................................................18

Model Execution ..............................................................................................................19

Conclusion ..........................................................................................................................20
Trademarks .........................................................................................................................20
APPENDIX ..........................................................................................................................21
MDD Buyers Checklist........................................................................................................21
Key Criteria for Vendors when looking at a UML based Model-Driven
Development Environment ..............................................................................................21
Model Driven Development Environment Buyers Checklist: Key Criteria when looking at
UML based Model-Driven Development Environments.......................................................22

Table of Figures

Figure 1 - Percentage of design completions according to schedule ....................................6

Figure 2 - Months of Delay for late completion or cancellation..............................................7

Figure 3 - Comparing Final Design Result to Pre-Design Expectations ................................7


Reducing Embedded Development with UML2.0
3 of 26

© 2004 Embedded Market Forecasters






















This page intentionally blank

Reducing Embedded Development with UML2.0
4 of 26

© 2004 Embedded Market Forecasters

Overview
Embedded designs have become increasingly more complex while the windows of
opportunity continue to shrink. In the global world of embedded enterprise, it is not
unusual to find OEMs and systems integrator design teams spread out by geography,
operating in different time zones, experiencing overlapping areas of responsibility and
finding bugs much too late in the design cycle.
The resulting design delays are expensive (EMF data shows a consistent 4-month
average delay with 56% of all embedded designs completely behind schedule). 11% of
embedded designs are cancelled and the average time-to-cancellation is nearly 5-
months. Engineering time associated with such delays and cancellations is very
expensive and does not include missed opportunity costs.
Annual surveys by Embedded Market Forecasters (EMF) of embedded developers has
clearly shown that software development is responsible for more than 80% of design
delays and associated design complications. This data also reports on embedded
developer responses to design complications. When asked how close their final design
was to pre-design expectations (for performance, systems functionality, features and
schedule) over 33% of respondents indicated that their final design was NOT within
50% of the pre-design expectation.
Respondents then indicated what steps they take if the final design is unacceptable.
The five most mentioned actions either involved extensive reengineering or removal of
systems features (see EMF report 2003: Embedded Hardware/Software Design
Preferences).
Whether the system is poorly conceived, specified or whether crucial algorithms fail to
adequately address systems performance, traditional methods of embedded software
development are yielding to a process known as Model-Driven Development (MDD).
MDD is used to more clearly define design specifications, test systems concepts and to
automatically generate code for rapid prototyping as well as for software development.
One of the major advances in software engineering design has been the use of the
Unified Modeling Language (UML) for enabling embedded design efficiencies.
Pioneered by companies IBM/Rational and I-Logix for embedded applications,
notwithstanding its value to addressing very complex designs and (with certain
commercial offerings) the ability to go from Statecharts to source code (automatic code
generation), the technology suffered due to incompatibilities of the various commercial
UML products. Despite the fact that there was a UML standard, vendors offered their
own extensions and variations that resulted in UML incompatibilities.
In order to rectify the problem of scalability and to make UML more available to OEMs,
embedded developers and systems integrators, the major vendors and users came
together to create the UML 2.0 standard.
From a commercial viewpoint, vendors that support the UML 2.0 standard need to find
ways to differentiate themselves from their competitors while maintaining their support.
In this report, data is presented to demonstrate that UML-based technologies, including
simulation-modeling, rapid prototyping, hardware-in-the-loop testing and automatic code
generation, offer better design results with considerable savings to OEMs, embedded
developers and systems integrators.
An analysis of how the major UML vendors differentiate themselves while maintaining
support for UML 2.0 is also presented.
Reducing Embedded Development with UML2.0
5 of 26

© 2004 Embedded Market Forecasters

Looking at Embedded Design Processes
The following information was developed from the Embedded Market Forecasters 2003
Survey of Embedded Developers (www.embedded-forecast.com). The results
presented are consistent from surveys conducted over the past 24 months.
Figure 1 presents a summary of design results according to schedule. The results are
also cross-tabbed according to architecture and vertical markets.
Percentage of Design Completions According to Schedule

Ahead Behind
Cancelled
Outsourced

Total Response

8-bit
15.5%

17.1%
54.0%

53.7%
13.1%

13.7%
11.6%

11.3%
16-bit 18.0%
49.5%
12.6%
12.5%
32-bit 15.4%
53.5%
13.1%
11.4%
64-bit 17.7%
47.3%
11.9%
16.8%
DSP 17.2%
54.8%
12.9%
11.3%
FPGA 15.8%
54.3%
12.6%
11.0%
Auto-Transport
14.8% 55.4% 15.1% 13.1%
Avionics
15.1% 52.0% 11.8% 15.2%
Bus Mach & Peripherals
15.1% 52.9% 14.8% 11.6%
Consumer Electronics
17.2% 52.3% 14.8% 11.9%
Datacom
11.8% 57.2% 16.5% 13.2%
Telecom
10.6% 60.2% 18.3% 7.5%
Electronic Instrumentation
18.3% 57.3% 13.3% 11.1%
Industrial Automation
19.1% 51.3% 13.0% 8.1%
Medical
18.1% 56.2% 11.6% 11.3%
Military
17.6% 52.1% 8.3% 14.3%
Figure 1 - Percentage of design completions according to schedule

The cost of del