Get ready, get set, race! Position and orientation data for autonomous navigation

GPS World, Sept, 2006 by William "Red" Whittaker, Louis Nastro

The DMI is another essential piece of the POS LV hardware used onboard both Sandstorm and Highlander. It outputs pulse data that represents fractional revolutions of the instrumented wheel. These pulses are converted by the POS LV into measurements of incremental distance traveled by the vehicle when no GPS is available. For both vehicles, the DMI helps bridge GPS outages and provide POS LV with incremental distance estimation, and acts as an input into the velocity controller for detecting when the vehicle may be stuck. Wheel slippage is monitored by comparing the DMI output to the velocity reported by the POS LV system. If, for example, the system reports speeds over 5 meters per second and a velocity of 0 meters per second, the vehicle will execute a set of protocols using the perception system and POS LV data to find the truth and make a string of control adjustments that will help find the way to the next pre-programmed point.

Autonomous Vehicles

Autonomously navigating at high speed for long distances across the desert, as was required in the 2004 and 2005 DARPA Grand Challenge events, necessitates robust and capable robots. In developing Sandstorm and H1lander, the Red Team opted-for proven, rugged designs to conquer the most difficult terrain conditions.

Built upon a gutted Humvee chassis, Sandstorm traveled farthest in the 2004 DARPA Grand Challenge, halting at mile 7.2. Sandstorm returned less than a year later to champion the pre-race tests and qualification rounds. The Sandstorm sensor platform uses a mechanical solution for easing the vibration placed on sensors caused by uneven terrain. With sensors on an internal plate, Highlander is built on a newer design intended to show how people and cargo can be moved comfortably.

Position Estimation

Autonomous vehicles sense, plan, and drive without the benefit of onboard human skill. Position estimation is essential for each of these. In a path-centric architecture, the fundamental action is to follow a path. A path data structure is pervasive through our approach. Pre-planned routes are provided to the navigation system and planning operations act as filters on the path. Positioning is used to steer sensor focus and allow the perception system to handle incompletely sensed terrain.

The Red Team's path-centric architecture provides a simple method for incorporating a pre-planned route. It reduces the search space for a planning algorithm from the square of the path length to linear in the path length, since planning is performed in a corridor around the pre-planned route. The path-centric approach avoids problems with arc-based arbitration such as discontinuities in steering commands (due to contradictory information) and jerky control (due to discrete arc-sets).

To use terrain evaluation data from multiple sources, the architecture uses a map-based data fusion approach. To provide this functionality, the architecture defines a second fundamental data type: the map. In this system, a map is a rectilinear grid aligned with the world coordinate system and centered on the robot. Each of the sensor-processing algorithms produces its output in the form of a cost map. Cost maps are a specific map type that represents the traversability of a cell with a numeric value (FIGURE 2).