IntroductionWhen a human operator drives a robot through agraphical user interface, he or she must have a propersituational awareness (SA), which has been considered inliterature one of the measures for evaluating an inter-face’s usefulness (Drury et al. 2007; Nielsen and Good-rich 2006; Olsen and Goodrich 2003; Scholtz et al.2004). The commonly accepted SA definition was givenby Endsley (1988) and adapted to HRI by Yanco et al.(2003). It distinguishes three components: human–robotSA, robot–human SA, and the human’s overall missionawareness. Within the human–robot awareness, twoaspects are important for the purposes of this paper:location awareness, defined as a map-based concept,which allows the user to locate the robot in the scenario,and surroundings awareness, which pertains to obstacleavoidance and allows the user to recognize the immediatesurroundings of the robot (Drury et al. 2007). Theseconcepts can be connected to cognitive distinctionbetween route knowledge and survey knowledge. Theroute perspective is closely linked to perceptual experi-ence: it occurs under the egocentric perspective in a‘‘retinomorphous reference system’’, that makes a personable to perceive oneself in the space (Herrmann 1996),with a special emphasis on spatial relations betweenobjects composing the scene an agent is situated in. Thisseems quite necessary for a correct surrounding aware-ness. In contrast, survey perspective is characterized byan external and allocentric perspective, such as an aerialor map-like view, allowing a direct access to the globalspatial layout (Cohen 1989) and giving information aboutstructures and global relations, but no clues about dis-tances between objects and information about turningpoints, perspective and visual appearance of landmarks,for example, and this can be connected with the locationawareness.Intra-scenario operator mobility is claimed to be a strongadvantage for acquiring SA within a robot teleoperation,only narrowed by portable devices’ limits. In contrast,remote controlling is supposed to give interface’s advan-tages (because it allows bigger screens and differentsources of information), but very bad SA.Our case study evaluates the performances of humanoperators remotely or directly driving a robot by meansof diverse human–robot interfaces. When the operator isnot physically in the navigation scenario, the interfacemust enhance his or her spatial cognitive abilities byoffering multilevel information about the environment(route and survey), and this can be done mainly withcomplex interfaces. In contrast, if the operator is insidethe scenario, part of the information can be acquired bydirect observation, depending on the visibility the oper-ator has, but the interface has to face the portable devi-ces’ limits.In our experiment, both laser and the map views, rep-resenting respectively ‘‘route’’ and ‘‘survey’’ knowledge(Siegel and White 1975; Tversky 1991), are available indesktop and PDA interfaces. The only (but crucial) dif-ference is that in the PDA the operator must change the