Goal-Based Domain Modeling as a Basis for Cross-Disciplinary Systems Engineering - Intentional Perspectives on Information Systems Engineering - page 84

• a fast requirements capture process using the domain model as starting point but

permitting project-specific extensions,

• a domain model-based similarity search identifying similar finalized projects as

sources for reuse,

• a means to enable domain model evolution, i.e.

- a mechanism to synchronize existing requirement models with a changing
domain model (in order to keep similarity search results accurate) and
- a mechanism to detect candidates for domain model extensions and reductions
from the project history, and last not least

• a transformation to subsequent modeling formalisms as a means to quality

assurance and a seamless integration into the overall development process.

In summary, our approach adopts a goal and dependency oriented modeling for-
malism suited to tackle interdisciplinarity and makes specific domain knowledge
available as concrete models. This really puts the domain knowledge at the cen-
ter of support. The suitability and effectiveness of the approach is exemplified by
applying it to control systems in the automotive domain.
The chapter is organized as follows. In Sect. 2 the field of control system devel-
opment is introduced to instantiate the abstract challenges for interdisciplinary,
project-driven SMEs for the requirements engineering phase of a combined soft-
ware and control systems development approach. In Sect. 3 the details of the domain
model based requirements engineering approach are presented. The role of the
domain model is elaborated especially with regard to how it helps SMEs to address
their particular problems. Section 4 discusses related work. Section 5 summarizes
the approach and outlines the chances for applying the developed solution to other
similar fields.

2 Case Study: Automotive Control Systems Development

Control system functionality, for example in cars, increases the comfort and safety
of driving a car or reduces the fuel consumption and exhaust gas emissions [ 1] .
The task of a controller is to continuously compare and adapt the current value(s) of
some system to some possibly changing desired value(s) [ 25] . Thereto the controller
interacts with the controlled system via sensors and actuators. Sensors measure
specific values of the controlled system and actuators change input variables of
it. Assume the controlled system to be a combustion engine. The controller has
to decide on the appropriate amount of fuel as well as the best point in time for
injection and ignition reflecting the user's demands (via the accelerator). A typical
sensor in this context is the knock sensor to detect self-ignitions that can damage
engine components. In this case the controller reacts with adapting the time for
ignition. Experiences and knowledge in physics, mathematics, and control theory
are required to design a stable controller with good performance.