The underlying question of this project, entitled Primate-inspired ground vehicle control, is: What role does mental rotation play in navigation if any and under what circumstances can it prove useful in support of robotic navigation?
This project was devised to address the ONR's FY11 Basic Research Challenge on Biologically-Inspired Flow Field Computation for Sensing and Control of Ground Vehicles. The research is being carried out in conjunction with the BORG Lab at Georgia Tech. The objectives of this project are:
- Understand, create, and apply the methods and models by which primates encode and use/manipulate spatial information (focusing on mental rotation).
- Develop perceptual techniques that allow for the encoding and manipulation of spatial information (BORG Lab).
- Integrate the ideas drawn from biology into a proven navigational framework and test the navigation algorithm generated using these ideas on a robotic platform.
This project has generated two different semi-reactive navigation algorithms that explore the mechanism of mental rotation and demonstrate its potential usefulness in navigation. Each of these algorithms uses a different encoding of the environment surrounding the agent. Given the successful creation of these algorithms, the lab is considering the potential usefulness of mental rotation in the context of advice-giving. If a person external to a robotic agent described the agent's goal location relative to some known "key" object(s) that is (are) within the agent's view, the agent can recognize this (these) object(s) and use an algorithm similar to the biologically-inspired algorithm that has already developed in order to navigate successfully to its goal location.