|
Project Summary | Autonomous
Tour Guide | Competitions
Project Summary [top]
Powerpoint Overview
(Right click and use "Save As")
Functional Block Diagram
(.PDF)
Webcam Tutorial (.PDF)
The Robot is driven by
two 1.75 HP DC motors at a pulse-width-modulation (PWM) frequency of 16
kHz, the 300 lb. robot utilizes a differential steering scheme and is
exceptionally maneuverable. To achieve the necessary level of autonomy,
software and hardware redundancy is employed in high and low-level
sensor arrays. Current low-level sensor techniques include a
combinational logic based collision detection system, ultrasonic
ranging for short-range proximity detection, and IR sensors for
drop-off detection. Low-level sensor data is first preprocessed by
slave microcontrollers, and then bussed to the onboard PC, whereas
high-level sensors communicate with the PC directly. The high-level
sensor array is presently comprised of a PNI TCM2-50 three-axis digital
compass module and a Trimble AgGPS 114 differential global positioning
system (DGPS).
Future additions to the
high-level array include a stereo vision system, and laser range
finder. Using a wireless VLAN, a remote laptop can access the onboard
computer and monitor all system parameters in real-time via a custom
GUI interface. Specific GUI display features include real-time obstacle
mapping, path-planning simulations, compass heading information,
collision and low battery alarms, plus plots of motor current and
vehicle velocity vs. time.
The versatile
WunderBot V will undoubtedly serve as a base platform for
advanced engineering projects in the future.
Autonomous Tour Guide
[top]
In addition to designing the WunderBot V platform for the Intelligent Ground Vehicle Competition, our team
has opted to impose several additional system constraints that will
allow the robot system to function as an autonomous tour guide. The
tour guide will function as follows:
1. The GPS will act as a
guiding device for WunderBot’s navigation as it provides campus tours
to prospective students. The robot will recognize specific GPS
coordinate locations and associate them with narratives about
particular campus locations (i.e. buildings, monuments, and sports
fields) stored in the robot’s memory.
2. A stereo vision system combined with a neural network will enable
the robot to recognize obstacles (i.e. people or debris) in its
predefined path, and adapt to them in real-time.
3. Internet users will
be permitted to control an on-board, motor mounted web cam, allowing a
360 degree view of the robot’s environment. This will allow prospective
students to receive campus tours via the World Wide Web.
Competitions
[top]
In 2008 , the IGVC will
celebrate the 16th consecutive year in which teams from around the
globe push the limits of machine intelligence. Sponsors such as DARPA,
The Department of Defense, GM, and the US Army provide generous
monetary awards to winning teams in all three of the competition
challenges:
1. Autonomous Vehicle
Competition – An autonomous ground vehicle must navigate through an
outdoor obstacle course within a prescribed amount of time and without
exceeding 5 mph.
2. Navigation Challenge
Competition – A vehicle must autonomously travel from a starting point
to a number of target destinations and then return to its initial
location given only the GPS coordinates of the targets.
3. Vehicle Design
Competition – A panel of expert judges will review a written report, an
oral presentation, and examine each vehicle in order to determine which
is the most innovative.
|
