Evaluation of Development Process and Methodology for Co-Models

Abstract

An embedded control system often requires a tight association between computational and physical system components. In such cases, embedded system development is difficult, as it requires the collaboration among stakeholders with different backgrounds (software engineers, mechanical engineers, managers etc.). With the constant increase in design complexity, caused by advances in implementation technologies, new ways of approaching embedded system development are needed.

This thesis presents an evaluation of a tool-oriented development process and methodology, supporting embedded system development. The philosophy of the development process and methodology, is that design complexity can be managed through collaborative work and multi-disciplinary modeling. To obtain input for the evaluation work, the development process is applied during a case study, involving the development of a route following robot and a model of this. To demonstrate the value of this model, it is simulated to predict route completion times for the physical robot.

The evaluation work identifies possibilities and challenges of the development process and methodology, with respect to traditional physical
prototyping. This will support developers in choosing the most optimal way of approaching development. In addition to this, suggestions for extensions to the methodology are provided. These intend to increase
the value the development process and methodology may bring the
development work.