Familiarity Navigation

Distributed Autonomous System Laboratory

Collaborators and Students: 

Alex Suhren (Student)

Dr. Christopher Crick (CS, OSU), Dr. Doug Gaffin (Biology, OU)

 

Project Description: 

In this project we are working on localization and navigation techiques for computationally constrained robots that do not require GPS. Bees and Ants navigate complex environments through the use of low-resolution vision alone. The familiarity hypotheses states that these animals navigate complex environments by moving towards familiar scenes, which are low-resolution representations of the environment. We are working on a robust navigation system that utilizes advanced machine vision techniques to build a localization and navigation framework inspired by the familiarity hypothesis. In our current research, a ground robot uses 360 degree field of vision camera to learn to navigate in cluttered environments without using GPS or odometry.