Linköping University: Students Alumni Trade and Industry/Society Internal Search

Student Theses

Listed on this page are the student theses that have been or are currently supervised at ESLAB. If you are interested in writing your thesis at ESLAB, please visit this page.

Optimization and Search in Model-Based Automotive SW/HW Development (in co-operation with Volvo Technology AB)

Contact person: Petru Eles


Models are increasingly used in the development of embedded HW/SW for the automotive industry. Several industry research projects (AUTOSAR, ATESST and ModelBus to name a few) have studied and proposed standardized languages, formalisms, data exchange formats, methodology, and platforms for automated tool chains for model-based development. The current state-of-the-art thus enables portions of the development flow to be fully automated, with little or no intervention by designers. There are still many design steps that are hard to automate or cannot be replaced by traditional algorithms. For example design steps that require artistic talent, require experience of the designers that is hard to formalize, or have high computational complexity.

In recent years the increasing computational power has enabled the use of computationally expensive search and optimization algorithms for many different engineering tasks. Machine learning, heuristic optimization and planning algorithms have been combined with guidance and goals from designers to better explore a design space and find optimal solutions. This project will apply these ideas in model-based development of embedded HW/SW in an automotive context. The overall goals are higher quality designs, shorter development time and time to market and lower development costs.

This is an initial project into an area with high potential to be a key component in future automotive development. It is a joint project combining the expertise of Volvo Technology AB, Chalmers University of Technology, and Linköping University.

The thesis students will be seated at Volvo Technology’s premises in Göthenburg.

Project Description

The aim of this project is to assess the feasibility of introducing algorithms and techniques for machine learning / search / heuristic optimization in a model-based development tool chain for automotive embedded systems.

  1. Study and describe the example case, i.e. the tool/model selected by Volvo.
  2. Select a few relevant optimization algorithms to be evaluated.
  3. Implement and apply the optimization algorithms.
  4. Evaluate and compare results with the existing (human designer-based) solution(s).


Students interested in this topic should preferably have knowledge, experience and interest in:

  • Software development
  • Search and optimization algorithms

Date Added: May 23, 2011

Completed student theses

2010 2009 2008 2007 2006
  • Karl Andersson SoC Test Data Compression in an Abort-on-fail Environment. LITH-IDA/DS-EX--06/010--SE

  • Karin Hedlund, Multi-site SOC Test with Replace-on-fail and Module Configuration. LITH-IDA/DS-EX--06/003--SE

  • Thomas Jansson, Core-based System-on-Chip Test Scheduling with Process Variations. LITH-IDA/DS-EX-ING--06/001--SE

  • Mikko Selkälä , Test Data Analysis for Accurate Power Estimation. LITH-IDA/DS-EX--06/004--SE, Produced publication
2005 2004 2003
  • Sriharsha Charan Reddy Naidu
    Automatic environment generation for formal verification of component-based systems

  • Jirong Zhu
    Towards an extensible visual editor for embedded systems representation models

  • Fabien Lambergeon
    C++ Analysis Library for Deterministic and Stochastic Petri Nets

  • Alexey Sinelnikov
    Testability Analysis and Enhancement for Self-Testable Digital Designs

  • Henrik Friman
    Petri Net Class Library and Translation from PRES+ to Timed Automata

  • Johannes Petersson
    A flexible simulator for control-dominated distributed real-time systems
2002 2001
  • Oskar Blomberg
    Modelling and simulation of real-time systems with discrete and continuous behavior
  • Daniel Karlsson
    A front-end to a Java-based environment for the design of embedded systems