Page 167 - Introduction to AI Robotics

150
Exercise 4.2 4 The Reactive Paradigm
Describe the difference between robot control using a horizontal decomposition and
a vertical decomposition.
Exercise 4.3
List the characteristics of a reactive robotic system.

Exercise 4.4
Describe the differences between two dominant methods for combining behaviors in
a reactive architecture, subsumption and potential field summation.
Exercise 4.5
Evaluate the subsumption architecture in terms of: support for modularity, niche
targetability, ease of portability to other domains, robustness.

Exercise 4.6
Evaluate potential field methodologies in terms of: support for modularity, niche
targetability, ease of portability to other domains, robustness.

Exercise 4.7
What is the difference between the way the term “internal state” was used in ethology
and the way “internal state” means in behavioral robotics?
Exercise 4.8
Diagram Level 2 in the subsumption example in terms of behaviors.
Exercise 4.9
When would an exponentially increasing repulsive field be preferable over a linear
increasing repulsive field?
Exercise 4.10
Suppose you were to construct a library of potential fields of the five primitives. What
parameters would you include as arguments to allow a user to customize the fields?
Exercise 4.11
Use a spreadsheet, such as Microsoft Excel, to compute various magnitude profiles.
Exercise 4.12
Return to Fig. 4.17. Plot the path of the robot if it started in the upper left corner.

Exercise 4.13
Consider the Khepera robot and its IR sensors with the RUNAWAY behavior instanti-
ated for each sensor as in the example in Fig. 4.19. What happens if an IR breaks and
always returns a range reading of N, meaning an obstacle is N cm away? What will
be the emergent behavior? and so on. Can a reactive robot notice that it is malfunc-
tioning? Why or why not?