We think we know that we will be able to navigate indoors using SLAM processing to keep us from running into walls and other obstacles. However it seems likely that this will not work outdoors, even if we constrain ourselves to a single campus. This project is an investigation into previous work in this arena as well as experimentation with what might work.
Notes and Steps
A way to think simply about this problem is to imagine that the robot is placed at outdoors at the entrance of building one and needs to navigate to the entrance of an adjacent building over a straight and flat path. What are all the ways in which this might be done?
- Investigate previous work in research and practice towards addressing this problem
- Test various ways to orient the robot in space while outdoors
* Some kind of triangulation of beacons (a la blue tooth)
* Some kind of machine vision of existing landmarks
* Or of artifical targets that are easy to identify visually
* Other ways
- Write a report with the findings