Abstract
Development and maintenance of virtual reality engines are coupled with large effort. It is therefore common today, to use existing solutions originating from the entertainment sector. This is often a compromise, since they fulfill individual requirements only in parts, due to their different background. The decision for a specific engine can have a large effect on the effort required to implement own functionality. The number of existing engines further complicates decision making. To enable a comprehensible and replicable decision making, we propose a structured selection process. In a multi-step approach, first the requirements and criteria for comparison are identified and analyzed. A pre-filtering is then used to select a feasible number of engines which are then compared in detail.