Video 1 shows our developed MUV exploring and mapping the accessible area of a whole building the structure of which is completely unknown. The aerial view of the building is shown in Figure 1. Figure 2 shows the map of the building and the trajectory of the MUV whereas Figure 3 shows the occupancy grid map (OGM) developed by the MUV. Figure 4 shows the OGM on the aerial view of the building. The results verify the two capabilities of our SLAM technology:
No need for loop closure: While many existing SLAM technologies require update of a map using loop closure due to the error accumulated during travel, our SLAM technology will need no correction using loop closure due to its high-accuracy mapping capability;
Accuracy: Accuracy we could achieve has been found with less than 5cm error for a building having an area of 80m x 65m.
Finally Figure 5 shows the three-dimensional (3-D) map of one of the rooms in the building created by our SLAM technology and with the addition of another LiDAR scanning vertically. Because of the high accuracy of our SLAM technology, the 3-D map created remains very accurate. It has been found that the 3-D map can be created in the order of 10cm.
Figure 1: Map of Hawkins Research Building
Video 1: Indoor SLAM (@ Hawkins Research Building)
Figure 2: Map of Hawkins Research Building and trajectory of MUV
