Fraunhofer FOKUS

MiLEST Toolbox
Model-in-the-Loop for Embedded System Testing

At the early stage of new system functionalities

development a model serves as a primary means for other libaries
product family
MiLEST
model-based testing
including the novelties, yet there is no code, no MATLAB
scripts callback
functions
hardware, and thus no real reference signals for
transformation
testing. In MiLEST FOKUS proposes a new method SIMULINK functions
simulation & model-based design

for the stimulation and evaluation of embedded

hybrid systems behavior which breaks down require- MATLAB
technical computation

ments into characteristics of specific signal features. A

novel understanding of a signal is defined that
The testing library is divided into four composite parts.
enables us to describe it in an abstract way based on
its properties e.g., decrease, constant, maximum. Model-based Testing

MiLEST Integration in the W-Model MiLEST Model-based Testing (MBT) is testing in which the
entire test specification is derived in whole or part from
MiLEST applies to the software built into embedded sys- both the requirements and an SUT model that describes
tems. In particular, it refers to the software models from selected functional aspects of the SUT. In this context, the
which systems are built. A model-based approach to func- term entire test specification covers both the abstract test
tional black-box testing is used to provide a test model. scenarios substantiated with concrete sets of test data
This is in marked contrast to current test methods that are and expected SUT outputs. It is organized in a set of test
generally dedicated solutions specialized for specific test- cases.
ing contexts.
Test Development Process
system &
acceptance test The Model-based Development paradigm assumes that
OEM
function design
system
vehicle integration
integration
vehic
icle
the System Under Test (SUT) model is already available
in th
he loop
op
abstract test hardware in the

architectural
specification
component
loop for integration and that the input/output interfaces are clearly defined
design integration
MiLEST
model-based testing and accessible.
MiLEST
model hardware model-based testing
SIMULINK in the loop integration
simulation and model-based design
hardware
software in the loop
design
SIMULINK MiLEST
simulation and model-based design Model-Based Testing

code software software verdicts analysis

generation integration in the loop

supplier
STEP 5 automatic execution

test control generation

The test method FOKUS proposes is realized in MATLAB/ STEP 4 manual refinement
Simulink/Stateflow environment.
test data generation
STEP 3 automatic generation
MiLEST Architecture
test specification
STEP 2 manual refinement
Technically, MiLEST is a Simulink add-on built on top of
the MATLAB engine that represents an extension towards test harness pattern application
model-based testing activities. MiLEST consists of a library STEP 1 automatic generation

including callback functions, transformation functions,

SUT as a model
and other scripts.
Fraunhofer FOKUS MiLEST Toolbox

Besides the analysis of the SUT specification, proper func- The structure of the test system consists of four different
tional, dynamic testing also requires systematic selection levels that can be built systematically. This makes it less
of test stimuli, appropriate test evaluation algorithms, error-prone while also leaving the test engineers plenty of
and obviously an execution or simulation environment. If scope for developing the complete test specification.
the above assumptions hold, a pattern for generation of
a test harness model can be applied on the SUT model as Features
denoted in step I. This is done automatically with a
MiLEST transformation function, giving a concrete frame Systematic, consistent functional test specification
for test specification. Further on, the test engineer refines
the test specification design (step II) using the concept of Signal feature-oriented paradigm
validation function patterns. Structures for test stimuli Graphical test design
and concrete test signals are then generated. This step
(step III) occurs automatically with application of the Test process automation
transformations. The test control design can be added systematic and automatic test data generation
automatically too (step IV). Finally, the tests may be exe-
cuted and test results obtained in the form of verdicts online automatic test evaluation
(step V). At the same time the quality of the test system Model-in-the-Loop test execution
specification produced is also assessed.
Reusable test patterns
Hierarchical Test System Abstract and concrete views

The MiLEST specifications reflect the structure of the sys-

tem requirements. The test development process and the
in-the-loop simulat
del- ion
abstractly aligned test system enable us to apply the con- mo

cepts of signal feature generation and detection mecha-

te
st
ev
nisms while also building test specifications systemati- test

alu rbitrat
ure sp
at
n

ec

a
ctio

a ti
fe
cally. In this context, the test system is understood as a

ifi

o n, n
abstra

al

ca
sign

tio

t e s t o r a cle,
hierarchically structured test model also called the test

io
n
MiLEST
design. method
t es

ion
tq

at
ua
lit m
y to
au

ss on
te te

test harness level test harness level ne rati

st

u
q

st alit
e
n
st y m ge ve
ls
ra
te etrics na acti
sig
gy test st re
te
abstraction

test requirement level

refinement

validation function level

Benefits

feature detection level

Testing in early design stages
Test of hybrid systems including temporal
Since the aim of the test system is not to provide the and logical dependencies
means for testing a single signal property but for validat- Traceability of test cases to the requirements
ing complete SUT models independently of their complex-
ity, structuring the test models in a proper way con- Traceability of verdicts to the root faults
tributes to the scalability and reusability of the solution. Increased test coverage and test completeness
Moreover, traceability of the test development artifacts is
possible and transformation potentials can be identified. Assured test quality of the test specification
Fraunhofer FOKUS, 2008

Contact
Dr. Justyna Zander-Nowicka Prof. Dr. Ina Schieferdecker
MOTION Competence Center Head of MOTION Competence Center
justyna.zander-nowicka@fokus.fraunhofer.de ina.schieferdecker@fokus.fraunhofer.de

Fraunhofer Institute for Open Telefon +49 30 3463 7000

Communication Systems FOKUS www.fokus.fraunhofer.de
Kaiserin-Augusta-Allee 31
10589 Berlin, Germany