A method and apparatus for evaluating capabilities of a system (e.g. an autonomous vehicle) with respect to the performance of a task, the method including: providing a first model corresponding to the task; providing a further model corresponding to the system; providing a plurality of mappings from one model to another model; determining whether there exists a transformation trace either (i) from the first model to the further model, or (ii) from the further model to the first model, wherein each transformation trace includes one or more of the specified mappings; and evaluating the capabilities of the system with respect to the task dependent upon the existence of the transformation trace.