6.186 | January IAP 2005 | Undergraduate

Mobile Autonomous Systems Laboratory



Lecture: Overview

Welcome, Staff, Logistics, Course Policies/Philosophy, Contest Preview: The Game, The Kit

Orc API and maslab.jar, OrcPad, PegBot, Soldering Tutorial


PegBot, OrcPad, Orc Hello World


Lecture, Part I: Mechanical

Lecture, Part II: Sensors and Cables


Cable Making: Have each team make its own IR sensor cable. Sensor Task: IR proximity with OrcPad feedback. Choose “bump/no bump” or edge-finder. Have other sensors available to play with. Evening: Java® for the clueless


Lecture, Part I: Software Engineering

On MIT Server and your robot (make, Ant, CVS), Design Principles, Threading in Java® (more mechanics than theory), Pitfalls

Lecture, Part II: Vision

Quick Review: Colors (HSV, Maslab colors, this year), Blue line (motivation), Determining color (thresholds on HSV, hysteresis, region support), See the tutorial!

Feature Detection: Methods (Template matching, Region growing, Clustering), Maslab features (What they are, Why they’re useful) How to locate them (corner detector, symbol decoder): Possible algorithm, Describe feature detection lab


Feature Detection Lab, Chassis Design, Strategy


Lecture: Mapping

Coordinate frames (image->robot->world), Image formation, ranging (image->robot, relative orientation), Odometry (robot->world, absolute orientation), Correspondence problem (we’ve solved it for you), Modeling uncertainty, Dealing with ambiguity, Possible algorithm


Work on Checkpoint 1


Lecture: Control


Simple Feedback (low-level control): Example: Korea-era smart bomb, Bang-bang control, Proportional control

State Machine (high-level control): Abstract definition and examples, Used for control of robot (6.004-style maze example, Brooks-style subsumption architecture), In Java® with threading: Leverage info covered in previous lecture, Full example, make code available


Checkpoint 1: Search Playing Field, Find Red Ball, Indicate on BotClient


Lecture, Part I: Advanced Vision

Clustering, Hough Transform, Stereo and Optical Flow, Statistical Models, Dealing with Noise, Performance

Lecture, Part II: Advanced Control

PID, Kalman, Whatever you want (maybe some high-level stuff?)


Mapping Activity, Purely Software Activity

7 Lecture: Design Review
8 Activity: Work on Checkpoint 2
9 Activity: Checkpoint 2: Score a Point
10 Activity: Mock Contest 1
11 Lecture: Programming 27 Robots: Distributed Algorithms for Robot Swarms and Engineering Creativity: Exercises for your Right Brain
12 Activity: Mock Contest 2
13 Lecture: Learning in Robots
14 Activity: Impounding
15 Activity: Contest Day