Meeting 2017-11-14 Test for localization

Definition of the test:

The rover (consistently describes by a point that is not necessarily present on the rover)localizes within a 4m x 4m grid such that the error (displacement) is not more than 10 cm (when the rover is static). No specific requirements are imposed on the method to localize. The rover will not be receptive to any explicit information about its distance i.e., the rover cannot subscribe to or request explicit displacement or pose information from a 3rd party.
3rd party – an agent that doesn’t reside on the robot or to which the robot is connected.
The rover can only be receptive to changes in the environment (such as terrain features or lighting conditions or silhouettes of itself or its own sensory data) throughout the entire process (that involves localization).
Location – pose within the test environment i.e, relative pose
The rover is operating on a 2D plane. Pose: [x, y, bearing].
(x,y) – > (m/ cm/ mm)
bearing – > (deg/ rad)
Scenario: 
Rover can be teleoperated
– rover drives along a path that is not more than 1 m wide on the inner side of the rectangle.
– The environment is a flat rectangular patch such that there exist no objects that occlude the sensor measurements
– Environment can be constructed such that no part of the same responds any different than how it did without the rover in the scene.
– 3rd party agents will not modify the calculated metric directly or indirectly. It is acceptable for the 3rd party agents to act like spectators in the scene.
Stretch Goal:
 
Perform loop closure experiment and determine the error.