Abstract: A method for controlling a construction robot in a working area of a construction site. The method includes detecting and determining at least one of a position, a geometry, or an orientation of the construction element within the working area; capturing a gesture of a user of the construction robot; generating a mixed-reality image which includes a representation of the construction element or at least of a part of the construction element and a virtual object at a virtual position within a virtual space, where the virtual space corresponds to the working area and the virtual position is chosen in dependence of the gesture; and making the construction robot manipulate the construction element at a position within the working area depending on the virtual position of the virtual object within the virtual space.