System Performance

Spring Performance Metrics

Requirement	Performance Measure	Result	Notes
CaBot avoids obstacles	CaBot stops or avoids obstacles that are no smaller than 2 ft. tall, 1 ft. wide, 0.5 ft. deep (placed on the ground) that are 4 meters or less from CaBot	CaBot avoided all obstacles, but sometimes took time to replan	A faster laptop would help decrease replanning time
CaBot detects pedestrians and alerts user	CaBot will announce the number of pedestrians in front of it at a range of 2 meters. CaBot will stop for stationary pedestrians less than 1 meter away and will decrease its speed for pedestrians with distance of 2 meters or greater.	CaBot identified pedestrians sucessfully, but gave some false positives and was sometimes slow to identify pedestrians	A faster laptop with a GPU would help decrease detection speed
CaBot localizes itself sucessfully	CaBot localizes itself with an error of 0.5 meters or less.	CaBot did it!	N/A
CaBot notifies the user of upcoming turns	While navigating, CaBot will notify the user of upcoming turns.	CaBot did it!	N/A

Requirement Description	Performance Measure	Result	Notes
CaBot starts navigation, leading one of the CaBot teammates with an iPhone	CaBot autonomously traverses through the waypoints established without running into a wall	CaBot occasionally ran into the walls during testing and FVE	The collisions with the walls were caused by sensitivity in the starting position, and a buildup of error in encoder measurements. We plan to fuse encoder data with LIDAR for localization to increase accuracy
CaBot's estimated position at each waypoint is accurate	Distance between CaBot and waypoint is less than 1.5 meters	During FVE, CaBot was less than 0.5 meters away from the waypoint	N/A
CaBot provides voice feedback to the user	CaBot alerts the user 2 meters before a turn and 0.5 meters after the turn. Error is +/- 0.5 meters	CaBot consistently met the performance measures in testing and FVE	We plan to add vibrational feedback to CaBot in the Spring