First Midterm (takehome)Your name:

1 2 3 4 5 6 7 8 9 10 total

10 20 20 25 10 25 15 10 10 20

Try to write the responses to the questions as if you were writing a conference paper. You must be precise. The concepts you use should be precisely defined. If you describe some concept such as decision tree, you have to illustrate it with examples and write all necessary mathematics. For instance, in case of quantum robot control you have to explain operators as unitary matrices, how to multiply them, etc.

In case of constructive induction you have to show example of decomposition. Etc. Avoid jargon. Sentences should be complete. I prefer the answers in Word than written by hand. Figures are very important to illustrate your point. References to literature and WWW other than my slides are very welcome.

In grading, I will take all these factors into account.

Robot Body versus Robot Brain versus application problemRobot Body versus Robot Brain versus application problem

Reactive – binary logic

Reactive fuzzy logic

Reactive quantum logic

Finite State Machine

Subsumption architecture

probabilistic GA GP Decision tree

grammar

Humanoid head

1. Dialog with a human, quarrel

2.Emotional monologue

3.

Facial gestures

4.Sport commentator

5. singer

6. Museum story teller

7, Receptionist 8. 9. 10.

Mobile robot

11. Obstacle avoidance

12. Territory mapping

13. Can collecting

14. soccer

15. sumo

16.

fencing

17. theatre

18.

dance

19.

Rescue mission

20. Synchronization tasks for a group

Walking Biped (KHR-1)

21. Walking reliably

23. All tasks as for mobile robot

24. Gymnastics, pushaps.

25. karate

26.Synchronization tasks for a group involving hands

27 28 29 30 31

Full body android puppet (does not walk)

31. All tasks as for humanoid head

32.3D obstacle avoidance

33.Do not hit himself

34 35 36 37 38 39 40

Problem 1. Try to complete the table. Write and explain briefly about each concept.

Problem 2. Select any combination of robot type, robot architecture and application and write in more detail about this type of robot. This should be not a learning robot. Define learning.

Problem 3. Select any combination of robot type and robot application but the architecture must involve some type of learning. Explain learning in much detail and illustrate with examples.

Problem 4. Favorite Robot

• Write short essay on your favorite robot:– Persona– Technology– Aesthetics– Mathematics (architecture, learning, behavior)

Problem 5. Morality and a Robot

• Write short essay on one moral aspect of using robots in future society.

• The robot is of your choice:– Military and Police– Anti-terrorist technologies– Entertainment– Surgery and medicine– Home helper– Advanced industrial robot versus unemployment– Any other

Problem 6. Crossover

• Crossover in Genetic Programming. You make take the rules from the book or develop your own ones.

• Show example trees of a Father that turns around, Mother that goes away and the child that goes to the gold in the labyrinth

• Show the labyrinth with obstacle, initial position and orientation of robots, location of gold.

• This robot (child) must be evolved only for the specific initial position and orientation and gold location.

Problem 7. Braitenberg Vehicles

• Design a Braitenberg Vehicle with memory• It has two emotional states, happy and unhappy.• When it is happy its behavior is “aggressive” (as

in my slides)• When the robot is unhappy its behavior is “shy”.• You decide what are the environmental

conditions to make robot happy or unhappy.

Problem 8. Lisp and GP

• Show how the Lisp program to count the number of elements in a list can be evolved step by step using GP (crossover and mutation).

Problem 9. Decision Trees and Probabilistic FSM

• Design (no software) a subsumption architecture for a soccer robot that uses decision trees and probabilistic state machines.

• Specify the tasks, the communication and how the machines and trees are used to select and execute actions.

Problem 10. Social behavior

• Explain and illustrate the behavior of a group of robots which are all either shy or aggressive robots.

• Each has a sensor in front and a sensor in back. White grid is zero, black grid is one. Another robot is black and one.

• Robots move one grid at a time.

• Red robots are aggressive and blue robots are shy.