Out line Robot ics Percept ion Robot ics Planning Reading: - - PDF document

1

Robot ics

J uly 26, 2005 CS 486/ 686 Univer sit y of Wat erloo

CS486/686 Lecture Slides (c) 2005 P. Poupart

2

Out line

• Robot ics

– Percept ion – Planning

• Reading: R&N Sect . 25.1-25.4

CS486/686 Lecture Slides (c) 2005 P. Poupart

3

Robot s

• Manipulat ors

– Physically anchored – Most indust rial robot s

• Assembly lines

CS486/686 Lecture Slides (c) 2005 P. Poupart

4

Robot s

• Mobile r obot s

– Shakey t he robot (SRI 1968)

• First mobile robot

– Service robot s (CMU’s Minerva)

• Museum t our guide robot

– Unmanned land vehicle (NavLab)

• Aut onomous highway driving

– Unmanned air vehicles

• Surveillance, crop-spraying, milit ary operat ions

– Aut onomous underwat er vehicles

• Deep see explorat ion

– Planet ary rovers (Nasa’s Soj ourner)

CS486/686 Lecture Slides (c) 2005 P. Poupart

5

Shakey t he robot

• First mobile robot

(SRI 1968)

CS486/686 Lecture Slides (c) 2005 P. Poupart

6

Nasa’s Soj ourner

• Planet ary

rover

2

CS486/686 Lecture Slides (c) 2005 P. Poupart

7

Honda’s P3 and Asimo

• Humanoid

robot s

CS486/686 Lecture Slides (c) 2005 P. Poupart

8

Robot Hardware

• Sensor s

– Range f inders (sonars, lasers) – Tact ile sensors – GPS – I maging (video cameras) – Propr iocept ive sensors (odomet ry) – Microphones

CS486/686 Lecture Slides (c) 2005 P. Poupart

9

Range scan example

CS486/686 Lecture Slides (c) 2005 P. Poupart

10

Robot ic Percept ion

• Challenge: noisy sensor s
• What st at ist ical model should we use t o

inf er t he st at e of t he world?

• HMMs (or DBNs)

s0 s1 s2 s3 s4

• 1
• 2
• 3
• 4

CS486/686 Lecture Slides (c) 2005 P. Poupart

11

Robot Localisat ion

• Sebast ian Thrun
• ht t p:/ / robot s.st anf ord.edu/ movies/ sca80a0.avi

CS486/686 Lecture Slides (c) 2005 P. Poupart

12

Simult aneous mapping and localisat ion

• ht t p:/ / robot s.st anf ord.edu/ movies/ mapping1-new.avi
• Sebast ian Thrun

3

CS486/686 Lecture Slides (c) 2005 P. Poupart

13

Robot Hardware

• Ef f ect or s

– Revolut e j oint s and prismat ic j oint s – Grippers – Wheels – Legs – Speakers

CS486/686 Lecture Slides (c) 2005 P. Poupart

14

Degrees of Freedom

• A robot has one degree of f reedom f or

each independent direct ion of movement

R R R P R R

6 degrees of freedom

CS486/686 Lecture Slides (c) 2005 P. Poupart

15

Degrees of Freedom

• How many degr ees of f reedom (DOF)

does a car have?

– 3 ef f ect ive DOF: x, y, orient at ion – 2 cont r ollable DOF

θ

(x,y)

CS486/686 Lecture Slides (c) 2005 P. Poupart

16

Degrees of Freedom

• Holonomic r obot s:

– # ef f ect ive DOF = # cont rollable DOF – Most robot arms – Easy t o cont rol – Complex mechanics

• Non-holonomic robot s

– # ef f ect ive DOF > # cont rollable DOF – Most mobile robot s – Harder t o cont rol – Simple mechanics

CS486/686 Lecture Slides (c) 2005 P. Poupart

17

Planning

• Challenge:

– Noisy sensors – Uncert ain act ion ef f ect s

• What st at ist ical model can we use?

– Part ially observable Markov decision process (POMDP) – Dynamic decision net work (DDN)

CS486/686 Lecture Slides (c) 2005 P. Poupart

18

POMDP (or DDN)

• Graphical Represent at ion

s0 s1 s2 s3 s4 a0 a1 a2 a3 r1 r2 r3 r4

• 1
• 2
• 3

4

CS486/686 Lecture Slides (c) 2005 P. Poupart

19

Greedy Pat ient Finding

• Sebast ian Thrun
• ht t p:/ / robot s.st anf ord.edu/ movies/ bad.people.avi

CS486/686 Lecture Slides (c) 2005 P. Poupart

20

POMDP P at ient Finding

• Sebast ian Thrun
• ht t p:/ / robot s.st anf ord.edu/ movies/ good.people.avi

CS486/686 Lecture Slides (c) 2005 P. Poupart

21

Ot her subf ields of Robot ics

• Mechanics

– Hardware engineering

• Cont rol

– Form of planning – Mainly concerned wit h “st abilit y”

CS486/686 Lecture Slides (c) 2005 P. Poupart

22

Next Class

• Next Class:
• Cour se wrap up
• Final exam inf o
• Ot her AI courses