Safe driving during the hours of darkness depends on the driver's visibility. A dynamic headlamp levelling system (HLL) keeps the range of the headlamps nearly constant for a maximum of visibility under given conditions without dazzling the oncoming traffic. Cost reduction and the further development of such systems up to advanced front lighting require special development methods. The model-based method is able to meet these requirements. In the case of a HLL, the advantages of a model-based approach, supported by a Hardware-in-the-Loop environment, can be used. Seen from a technical point of view, the advantage of the model-based development in comparison with driving tests is based in the reproducibility. Furthermore, the boundary conditions within the laboratory are controllable in comparison with driving tests that depend on the weather conditions, road conditions and the traffic situation. This paper analyses the requirements on future HLL with respect to cost reduction, reducing the development time and avoiding dangerous driving tests. It describes the model-based development within the simulation and the used Hardware-in-the-Loop test bench. Further on, an alternative concept is investigated and some results for this concept compared to an already existing system are shown. This paper described how costs could be reduced in the case of further development of a dynamic HLL. Being independent from driving tests means being independent from boundary conditions like weather or lightness. The contribution showed how the results of an alternative concept could be compared with a conventional concept. The test bench itself gives a first feedback about the results and especially the diagrams help to evaluate the concepts. Up to now there is no curve that indicates the minimal quality for the height h. Future works should investigate the height curves of several driving tests, so that a curve with minimal quality can be defined.