Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
144 views

Robotics Midterm Review

The document provides an overview of the topics and format for the EGN 3060: Introduction to Robotics midterm exam. It will cover A* planning, lab report comments, exam format which allows one page of notes, and key topics including robotics architectures, planning, and computer vision up to Lecture 13. Example questions are provided for A* planning, reactive architectures, and hybrid architecture modules.

Uploaded by

Grayson Savage
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
144 views

Robotics Midterm Review

The document provides an overview of the topics and format for the EGN 3060: Introduction to Robotics midterm exam. It will cover A* planning, lab report comments, exam format which allows one page of notes, and key topics including robotics architectures, planning, and computer vision up to Lecture 13. Example questions are provided for A* planning, reactive architectures, and hybrid architecture modules.

Uploaded by

Grayson Savage
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 12

EGN 3060: Introduction to Robotics

Midterm Exam Review Instructor: Dr. Gita Sukthankar Email: gitars@eecs.ucf.edu

A* Review
Can calculate f,g,h for every node on the map assuming 4 point connectivity H (heuristic estimate of distance to goal)

G (cost of traveling from start state)

F is the priority of the node; the order that it gets pulled off the queue

F=g+h

Lab Report Comments


Main issue with lab reports were people who omitted their cost maps or pictures of the GUI Cost map update order matters:
9 10 11 12 8 1 1 5 7 1 5 4 6 5 1 3 5 4 3 2 (goal)

Exam Format
Questions: T/F, multiple choice, short answer, simple problems (no calculator needed) Can bring into the exam:
1 sheet of notes (front and back covered)

Topics
Overview on Robotics Architecture* Planning* Computer Vision (up to Lecture 13)* Lab Material and Homework Questions Book chapters: 2, 4, 5, 7, and 9

Overview on Robotics
Perception/cognition/action Representation/world model/frame problem Map representations: topological, metric General concepts related to autonomy and teleoperation but not the specific paradigms Interaction between planning/vision/localization Dont need to know: example robotic or vision applications described in class This is less important than the other topics
6

Architectures
Types of architectures Deliberative, reactive, behavior-based, hybrid STRIPS planning and means-end analysis Reactive: Subsumption Potential fields Hybrid: example module types Know the strengths and weaknesses of the different paradigms Be able to draw a finite state automata Dont need to know: names of architectures other than the ones above
7

Example Questions
True or false: Reactive architectures need to maintain a persistent model of the world state (the state of entities in the world) to function effectively. Vector summation is used as the basis for combining behaviors in which type of architecture: a) deliberative b) potential fields c) hybrid d) subsumption e) all of the above
8

Example Question
Hybrid architectures often contain the following modules: mission planner, cartographer, sequencer agent, behavioral manager, performance monitor. Describe and provide an example of tasks handled by three of the modules.

Cartographer

AuRA Architectural Layout Performance


Monitoring Mission Planner

Sequencer Behavioral manager (mgr+schemas)

Emergent behavior

Planning
Dockworker robot example General operation of the state-space planners (ideas like the Sussman anomaly) Search-based planning: A* Wavefront planning Visibility and Voronoi graphs Spatial decompositions Sampling-based planners Understand ideas behind: Graphplan, hierarchical task network planners, plan repair (D*), sensor-based planning Dont need to know: configuration space, complexity 11 results for planning

Image Processing
What makes computer vision difficult Color models: RGB and HSI Thresholding Histogram methods Morphological filters: erosion and dilation Linear filters: what are the results of various linear filters Simple cross-correlation and convolution examples Visual servoing and PID control Stop at the end of Lecture 13

You might also like