According to Veterans Affairs staff, up to 50% of a nurse's day is spent transporting patient care items. To aid nurses, fully autonomous robotic transporters driven by Artificial Intelligence (AI) could access the entire hospital and make simple deliveries for nurses, giving them more time to deliver patient care.
I researched current AI pathfinding methodologies and found that the leading pathfinding algorithm, A*, doesn't utilize re-alignment opportunities needed in a hospital setting where a robot is more likely to go off course. I wondered... can current AI pathfinding methodology that is used for autonomous robots be improved to increase the accuracy of the robot reaching its destination?
I developed a pathfinding algorithm to create paths that maneuver the transporter around known obstacles, while considering accuracy and efficiency. I then tested the accuracy of the resulting paths of A* and my algorithm by measuring the accumulated variance of the robot's ending position from its target destination in feet. I used the following experimental variables:
Independent Variables: Path plotted by the AI algorithms used
Control group: A-star algorithm
Experimental group: Accuracy/efficiency balancing algorithm
Dependent Variable: Accumulated variance of the robot's ending position from its target destination measured in feet.
I plotted and measured the results from my experiment and found that if If the robot's algorithm is modified to incorporate aspects of its surrounding environment to correct its course, then the variance of the ending position of the robot from its target destination will be reduced by 36.5%.