Automated Migration of Legacy Functions and Algorithms to Model Based Design 2008-01-0747
Automotive companies have invested a fortune over the last three decades developing real-time embedded control strategies and software to achieve desired functions and performance attributes. Over time, these control algorithms have matured and achieved optimum behavior. The companies have vast repositories of embedded software for a variety of control features that have been validated and deployed for production. These software functions can be reused with minimal modifications for future applications.
The companies are also constantly looking for new ways to improve the productivity of the development process that may translate into lower development costs, higher quality and faster time-to-market. All companies are currently embracing Model Based Design (MBD) tools to help achieve the gains in productivity. The most cost effective approach would be to reuse the available legacy software for carry-over features while developing new features with the new MBD tools.
Ford Motor Company has successfully integrated the legacy Escape Hybrid control strategies into a model. The new model is now being used for development of the next generation of hybrid vehicle controls. This migration process was time consuming and resource intensive. Using the lessons learned from the mainly manual process, the conversion process has been optimized to dramatically speed up the integration of legacy software to support the migration of other legacy projects to MBD. This paper presents the process, the tools and the results of these efforts.