Adaptive controllers, based on backstepping, are verified on practicability, and are used to improve the control of an experimental setup: an XY-table. Theoretical bounds for the adaptive schemes indicate strong transient performance. So far little is known about the robustness of the controllers. The robustness, with state and output feedback controllers, is verified with experiments for the XY-table.The controllers based on state teedback showed good results. They were easy to implement and their parameter tuning appeared straightforward. The controllers based on output feedback were more difficult to implement. In experiments the range of possible tunings was restricted due to a conflict between numerical accuracy and real time requirements. Two extensions on the original control scheme are proposed. Both extensions result in more freedom in the controller design. In the allowable tuning range, the results were still inferior to the results obtained with state feedback.