Map designs

The flexibility of a computer-based map allows the inclusion of various features such as auto rotation of the map according to the user’s orientation, updating the user’s location in real time or providing users with interaction such as the ability to zoom in/out of the map, all of which are not available in a physical map. However, a careful approach needs to be taken when designing and developing the maps to ensure that they provide users with the required survey knowledge instead of confusing users. A well-developed map is useful for navigation and also spatial knowledge acquisition, which is crucial in many scenarios such as evacuation planning or looking for a particular location in a building (Li & Giudice, 2012). Adding unnecessary aids results in users having cognitive overload (Haik et al., 2002). Unnecessary aids could include adding a map and arrow, when the map is already sufficient to aid the users with the navigation.

An early approach to developing a map for a large scale VE is the 2D map, was based on the map principles of Darken and Sibert (1996). They suggest:

 Dividing the large scale VE into smaller parts that are shown on the map together with all the elements of the environment such as the paths, landmarks, etc.

 Showing the user’s position on the map.

 Orienting the map with respect to the direction in which the user is facing in the VE.

The first point, dividing the large scale VE into smaller parts, was done using a grid placed on the map (Figure 2-3). The results of the study demonstrate that this approach helps users to gain more accurate survey knowledge than using the map without any grid. The last point, orienting the map with respect to the user’s direction, could be substituted by using a cue that shows the orientation of the user rather than orienting the map, as suggested in a more recent study by Darken and Peterson (2001). This was also supported by the fact that users were not keen to rotate the map (although given an option to do so) when the map is provided with an indicator showing the direction in which the user is looking (Chittaro & Venkataraman, 2006).

A more interactive map was developed by Stoakley, Conway & Pausch (1995). Known as ‘World in Miniature’ (WIM), this interactive, immersive 3D map replicates a life-sized VE. The user’s interaction with the WIM is done using a tennis ball installed with buttons for interaction (e.g., moving an object) while the other hand holds a clipboard where the user can control the display of the WIM. For example, if the clipboard is raised, the WIM in the screen also rises. Any interaction in the WIM (e.g., moving an object) corresponds to the movement in the VE. The users view of the WIM through the Head Mounted Display (HMD). A drawback of WIM is that it displays the entire VE in a single map, which causes a problem for a large scale VE, because all details are cramped into the single map. This drawback was overcome by adding a scale and scrolling function to the WIM, allowing a detailed view of the environment (Wingrave, Haciahmetoglu, & Bowman, 2006). Using the scrolling function, (similar to a zoom in/out function) to allow a detailed view of a map also aligned with the suggestion by Haik et al. (2002). The zooming function of a map was preferred by the users who felt it supported the navigation (Hornb, Bederson, & Plaisant, 2002).

Another approach that could be used to overcome the crowded large scale VE is to divide the large scale VE into smaller parts, as suggested in the map principles (Darken & Sibert, 1996) discussed earlier. However, dividing a large scale into a small VE may not meet the objective of the WIM, which is intended to display the whole VE at once.

The early development of maps was more focused on a single level space, be it an open space such as a large sea or closed space such as building. Little is known about studies related to navigation in multilevel buildings. Navigation in multilevel buildings is often a problem since people have a tendency to lose their orientation when moving to a different

floor (vertical movement) and assume the layout of each level in the building is similar (Soeda, Kushiyama, & Ohno, 1997). This navigation is more complex because it requires both horizontal and vertical knowledge of the environment; therefore, presentation of a map needs to consider both horizontal and vertical aspects.

The lack of studies related to maps for multilevel buildings has been pointed out by Chittaro and Venkataraman (2006), who developed an interactive 3D Break Away Map (I3BAM), which was intended for multilevel buildings (Chittaro et al., 2005). The I3BAM was based on WIM, with some modifications made to suit multilevel buildings.

In I3BAM, users can only travel in the VE and the user’s current position and orientation in the VE are updated correspondingly on the map using an indicator (a sphere with an inverted triangle), similar to the suggestions made by Darken and Peterson (2001). The map only shows the floor where the user currently is (Figure 2-4 (a)).

Apart from using it while navigating the VE, the user could also use I3BAM to examine the layout of a building, by selecting the respective floors the user wishes to view. In WIM, displaying more than a single floor may block the view of other floors, particularly floors below the current view. In I3BAM, a user can select the floor s(he) wishes to view and the other floors are slid away, as shown in Figure 2-4 (b). A preliminary study using the I3BAM as a means to examine a physical building shows promising results, where the users were able to navigate in the real building based on the information presented in I3BAM.

Figure 2-4 The map with an indicator showing the user’s location and position at the current floor (a). The map is showing level 2, where levels 3 and 4 are slid to the left (b).

Figure 2-5 The 3D map (a) and the 2D map (b) of a multilevel buiding. The 2D map and the corresponding VE in a multilevel building (c) (Chittaro & Venkataraman, 2006).

In a more recent study, Chittaro and Venkataraman (2006) compare the effectiveness of the I3BAM by simplifying it into a 2D (Figure 2-5 (b)) and a 3D map (Figure 2-5 (a)). The 2D and 3D maps have common features in that both maps display all three levels of the building including the stairs where users can move between levels. Similar to the I3BAM, travelling can only be done in the VE, and the position and direction of the user in the VE is updated correspondingly on the map. Only the map with the user’s current position is displayed in detail whereas the remaining floors have only their respective borders outlined. Sliders are provided for the maps, where the 2D map could be rotated about the axis normal to the panel and the 3D map could be rotated around the horizontal and vertical axes.

Testing tasks included a way-finding task and a direct estimation task (pointing to an object in the VE from a predetermined location). The results of the study indicate that the 2D map assists users in performing the way-finding task efficiently (looking for different objects in the VE). No significant differences were found in terms of the acquisition of spatial vertical and horizontal knowledge.

Another study using a map as a navigation aid in a multilevel building is by Luo, Luo, Wickens and Chen (2010) who compare different conditions of an exocentric view: ‘3D floor map’, ‘3D building map’ and ‘transparent condition’. Similar to the study by Chittaro and

Venkataraman (2006), users can only travel within the VE and the map updates each user’s position accordingly. Movement between floors is done using the elevator. The conditions in the user study are shown in Figure 2-6.

In the 3D floor map, only the floor where the user is currently located is displayed, similar to the 3D map by Chittaro and Venkataraman (2006). In the 3D building map, a view of all floors is displayed at the same time with a transparent floor, which allows the user to view all floors regardless of which floor the user is currently on. In the transparent condition, no map is made available. However, the walls are transparent so that when the users move from one floor to another through the elevator, they gain an exocentric view of each floor (through the transparent wall) before they walk out of the elevator. Therefore, the users gain a larger overview from an exocentric perspective of each floor than that shown on the 3D floor map or 3D building map.

Figure 2-6 The 3D floor map, 3D building map and the transparent condition of a multilevel building (Luo, Luo, Wickens, et al., 2010).

3D floor map 3D building map

Figure 2-7 The The cut-away map according to the user’s viewing direction in the VE (Andujar et al., 2010).

The study included the judgement of relative direction, where the users need to point to an object in the VE from a given imaginary position and orientation. The results indicate that, in general, users of the 3D building map (where all levels of the building are displayed concurrently with transparent floors) achieve the most accurate direction judgement. The 3D building map is better in conveying vertical information (vertical relations between the building levels) compared to the 3D floor map.

Another study on multilevel buildings was conducted in a more immersive environment. Andujar, Argelaguet and Trueba (2010) developed an interactive cut-away map, an extension of WIM, to allow users to view objects that have been blocked by other objects (e.g., a wall), as shown in Figure 2-7. The results demonstrate that this map assists users with performing a search task more efficiently (searching for an object in one location and moving it to another location).