The Navigation technology in ERSP 3.1 provides various functions that enable a robot to be fully autonomous and move safely in its environment all with a standard web camera as its primary sensor. Specifically, with ERSP Navigation a robot can:
- move in its environment with knowledge of its location
- build a map of its environment
- plan efficient paths to reach destinations
- locate and dock into stations to recharge its battery
- avoid obstacles and hazards
The Navigation APIs provide access to functionality for controlling the movement of the robot including modules for teleoperation, obstacle avoidance, target following, localization, mapping, and path planning.
With vSLAM, a robot can operate in a previously unknown environment by using visual measurements to build a map and then later, to localize itself by recognizing visual features using that map.
A vSLAM generated map consists of a set of unique landmarks that the robot generates as it explores its environment. Because of its sophisticated method of processing sensor data and its use of patent-pending ERSP Vision, vSLAM functionality is robust to changes in lighting, moving people, and movement of objects.
vSLAM provides the first-ever, truly robust solution to robot kidnapping.
Path planning can be accomplished in two ways. A user can request a path and the robot will select one using waypoints and knowledge of obstacles it has collected during its time wandering in its environment. Path planning can be optimized if used in conjunction with vSLAM.
Obstacle avoidance functionality allows the robot to continuously detect obstacles and compute local paths to avoid them. Cliff/stair avoidance functionality provides a reflexive response to a hazardous situation to insure the robot's safe movement.
The exploration feature enables the robot to wander in an unknown environment safely. In conjunction with vSLAM and obstacle and cliff/stair avoidance, it provides a powerful method of autonomous map building.
The Occupancy Grid represents a two-dimensional map of the robot's environment using a fine-grained grid that models the occupied and freespace of the environment. The Occupancy Grid is useful for remotely controlling a robot, providing a live, as well as historical picture of obstacles and pathways. The robot also uses the Occupancy Grid to keep a historical picture of known obstacles on existing vSLAM maps.
Back to top
|
Overview
Navigation
Evolution Robotics stands behind every robotic product that has been "Built on Evolution."
