Model-driven development claims to offer the same improvement to developers that the step from assembly to procedural programming languages offered developers in the last century. In this paper we report on a project that assessed the feasibility of applying a model-driven development approach to the evolution of a non-trivial legacy system in an industrial setting. The project focused on whether the promises of model-driven development hold in the real world and whether it possible to adopt model-driven development without losing the investment in existing third generation language code. An existing component from a system was redeveloped and integrated back into the system using a current model-driven development tool at Invensys Rail Systems Australia. From experiences throughout the project, and with the tool, a set of requirements for future model-driven development and model-driven development tools was produced. These requirements are designed to both aid tool developers, and to give companies a way of assessing the maturity of both future model-driven development tools and model-driven development itself.