Regular Meeting Minutes
Apr 7, 2007, 10:30AM
Line Maze Challenge %GĀ‚Äď%@ Monty Goodson
í°»Manhattan Project Secretsí°… Jeff Steidl
í°»Reflections on Reflectionsí°… -- Mark Medonis
PDXBot Update %GĀ‚Äď%@ Larry Geib.
News & Reports
See PDXBot í°«07 update below.
Monty Goodson brought a new maze to the meeting. This meeting is the second-to-last in the table top challenge series. It includes multiple turns and dead-ends, as well as a goal circle. The final maze will include a loop.
Monty Goodson demonstrated his robot, which completed the maze successfully, and discussed his robot construction and strategy.
Mark Moore discussed a simulation program for his line maze solving strategy. The simulation represents the robot and the status of its sensors as it negotiates the maze.
Jeffí°«s maze solver will incorporate 16 analog line sensors (built of discrete LEDí°«s and photo diodes), positioned such that the line is always visible to a sensor. The design includes an FPGA and a microcontroller board. Jeff plans to use line sensor data and floating-point processing to calculate his robotí°«s velocity in a direction perpendicular to the line. This will allow the robot to correct for course errors more smoothly. The design also includes brushless motors %GĀ‚Äď%@ the generation of the drive signals for the motors will provide the robotí°«s odometry.
Jeff is also planning to design his own processor for this robot. His design is It will be a memory-to-memory architecture meant to handle variable data sizes from 1 to 128bits. The processor design supports a numeric encoding format to simplify programming with mixed signed and unsigned values. Floating point calculations will be emulated.
Mark works at a company that makes infrared thermometers used in silicon wafer processing.
Reflectivity is a physical processor that is common to the study of optics and heat transfer. Specular reflection is the well-known í°»bounceí°… of a light off of a reflective surface. Lambertion is the í°»scatteringí°… of light when it hits a surface. Perfect Lambertion would result in an even dispersion of reflected light in all directions.
When building a line sensor, you are basically trying to measure the amount of light reflected from a target surface. Doing this implies knowing how much light you are emitting, as well as the properties of the surface. Mark recommends using a constant current source when driving an LED to improve the stability of the light source in a reflective sensor. Ambient light should be controlled using optical filters or physical shielding.
Mark recommends Omron EE-SB5 for robotics line sensors.
Mark uses an LM339 circuit to condition IR sensor inputs before sending them to a digital input on a microcontroller.
PDXBOT will be held on May 19 at the Portland Doubletree Inn at Lloyd Center. RoboMagellan will be held on May20 in the PSU park blocks.
The next planning meeting is on Thursday April 12, at 7PM; Lucky Labrador 1945 NW Quimby. We are calling for help in the following areas:
Audio & Visual presentation
Contest trophy production.
Tech demos. %GĀ‚Äď%@ Contact Loren McConnell with ideas.
For RoboMagellan, we are planning on having í°»hardí°…, í°»mediumí°…, and í°»easyí°… cones. Touching the intermediate cones is not compulsory.
The Portland Dorkbots club will host a new competition for robots that make art.
Please visit www.pdxbot.org for event lists, rules, and updates.