Cal Poly Pomona

1
Cal Poly Pomona

• Robot Navigation
• Salomón Oldak, Ph.D.
• Electrical and Computer Engineering
• 2/8/06

2
DARPA Grand Challenge

• Congress mandate to increase the use of ground
  unmanned vehicles.
• Congressional Goal 1/3 of armed forces combat
  vehicles unmanned by 2015.
• Vehicles must be fully autonomous
• Nearly 250 miles on and off road in 10 hours.

3
First DARPA Challenge 2004

• First Challenge took place on 3/14/04
• Between Los Angeles and Las Vegas
• 1 Million Prize
• 25 Teams Participated
• CalTech, University of Florida, University of
  Alaska, Virginia Tech and others
• Palos Verdes High School

4
Ghostrider(Blue Team, Berkeley)
5
Blue Team Movie
6
Obstacles

• Paved Roads
• Overpasses
• Straight-Winding Roads
• Sand, Rock
• Underpasses
• Water
• Natural Obstruction

7
1st Challenge Results
Failure No Team Completed Task. Max Distance
7.4Mi
8
2nd DARPA Grand Challenge (10/8/2005)

• Success! 3 Robots Completed task!

9
The Winner Stanley
10
Stanford University (Sebastian Thrun)
11
Sensors Used

• GPS Antenna
• Laser Range Finder (Lidar) (30m)
• Video Camera (80m)
• Odometry (Photo Sensor on Wheel)

12
Algorithm

• Problems Vibration would trick sensors to
  imagine ghost obstacles, The vehicle thought
  its own shadow is an obstacle.
• Solution Teach the car. Assess weights to pixels
  as a human driver operates the car.

13
Future

• 43000 people die in traffic accidents/year in the
  US
• Robot driven cars will reduce of fatalities
• Accidents can be avoided
• Liability?

14
Spirit and Opportunity
15
Navigation

• Rovers are mostly teleoperated
• 2 stereo hazard avoidance cameras front and back
• 1 Stereo Navigation camera on mast
• Rovers moves 0.5m at max speed of 34m/h 0.02MPH

16
Robot Paradigms

• Paradigm
• A "view" of how things work in the world.
• a set of rules and regulations
• Paradigm Primitives
• SENSE, PLAN, ACT

17
Paradigms

• Hierarchical
• 1967-1990
• Reactive
• 1988-1992
• Hybrid Deliberative/Reactive
• 1990-

18
Hierarchical Paradigm

• The robot operates in a top-down fashion, heavy
  on planning.
• The robot senses the world, plans the next
  action, acts at each step the robot explicitly
  plans the next move.
• All the sensing data tends to be gathered into
  one global world model.

19
Shakey (SRI)
First AI Robot (1967-70)
20
Assumptions

• Close World World Model Contains everything the
  Robot needs to know
• Frame Problem The real-world situation is
  computationally feasible.

21
Problem

• Robots designed under Hierarchical Paradigm were
  VERY slow.

22
Reactive Paradigm

• Sense-act type of organization.
• The robot has multiple instances of Sense-Act
  couplings.
• These couplings are concurrent processes, called
  behaviours, which take the local sensing data and
  compute the best action to take independently of
  what the other processes are doing.
• The robot will do a combination of behaviours.

23
Biological Foundation of Reactive Paradigm

• Reactive Paradigm is based on observations of
  ethologists (study of animal behavior) and
  cognitive psychology (how humans think and
  represent knowledge)
• Biology provides existence proofs.

24
Behavior Definition (graphical)
BEHAVIOR
Pattern of Motor Actions
Sensory Input
25
Arctic Terns

• Arctic terns live in Arctic (black, white, gray
  environment, some grass) but adults have a red
  spot on beak
• When hungry, baby pecks at parents beak, who
  regurgitates food for baby to eat
• How does it know its parent?
• It doesnt, it just goes for the largest red spot
  in its field of view (e.g., ethology grad student
  with construction paper)
• Only red thing should be an adult tern
• Closer large red

26
Types of Behaviors

• Reflexive
• stimulus-response, often abbreviated S-R (like
  knee tapped). Hardwired
• Reactive
• learned or muscle memory. (Riding a bike,
  skiing, etc.)
• Conscious
• deliberately stringing together (Building a
  Robot)

WARNING Overloaded terms Roboticists often use
reactive behavior to mean purely reflexive, And
refer to reactive behaviors as skills
27
Example Cockroach Hide

• light goes on, the cockroach turns and runs
• when it gets to a wall, it follows it
• when it finds a hiding place (thigmotrophic),
  goes in and faces outward
• waits until not scared, then comes out
• even if the lights are turned back off earlier

28
Behaviors are Concurrent
29
What happens when theres a conflict from
concurrent behaviors?

• Equilbrium
• Feeding squirrels-feed, flee hesitate in-between
• Dominance
• Sleepy, hungry either sleep or eat
• Cancellation
• Sticklebacks defend, attack build a nest

?
30
Reactive Robots
RELEASER
behavior
SENSE
ACT

• Most apps are programmed with this paradigm
• Biologically based
• Behaviors (independent processes), released by
  perceptual or internal events (state)
• No world models or long term memory
• Highly modular, generic
• Overall behavior emerges

31
Example My Real Baby

• Behaviors?
• Touch-gt Awake
• Upside down Awake-gt Cry
• Awake Hungry -gt Cry
• Awake Lonely -gt Cry
• Note can get crying from multiple behaviors
• Note internal state (countdown timer on Lonely)

32
Subsumption ArchitectureRodney Brooks
From http//www.spe.sony.com/classics/fastcheap/in
dex.html
33
Runaway
34
Example Perception Polar Plot
if sensing is ego-centric, can often eliminate
need for memory, representation

• Plot is ego-centric
• Plot is distributed (available to whatever wants
  to use it)
• Although it is a representation in the sense of
  being a data structure, there is no memory
  (contains latest information) and no reasoning
  (2-3 means a wall)

35
Potential Fields (Example Navigation)


7
6
5
4
3
2
1
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 1 1 1 1 1 1 1 1 0 0 0 0
0 0 0 0 1 1 1 1 1 1 1 1 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2
36
Wavefront Algorithm


7
6
5
4
3
2
1
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
18 17 16 15 14 13 12 11 10 9 9 9 9 9 9 9
17 17 16 15 14 13 12 11 10 9 8 8 8 8 8 8
17 16 16 15 14 13 12 11 10 9 8 7 7 7 7 7
17 16 15 15 1 1 1 1 1 1 1 1 6 6 6 6
17 16 15 14 1 1 1 1 1 1 1 1 5 5 5 5
17 16 15 14 13 12 11 10 9 8 7 6 5 4 4 4
17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 3
17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2
(0,7) -gt (1,7) -gt (2,7) -gt (3,7) -gt (4,7) -gt
(5,7) -gt (6,7) -gt (7,7) -gt (8,7) -gt (9,7) -gt
(10,7) -gt (10,6) -gt (11,6) -gt (11,5) -gt (12,5) -gt
(12,4) -gt (12,3) -gt (13,3) -gt (13,2) -gt (14,2) -gt
(14,1) -gt (15,1) -gt (15,0)
37
Hybrid Deliberate/Reactive Paradigm

• The robot first plans (deliberates) how to best
  decompose a task into subtasks (also called
  mission planning) and then what are the
  suitable behaviours to accomplish each subtask.
• Then the behaviours starts executing as per the
  Reactive Paradigm.
• Sensing organization is also a mixture of
  Hierarchical and Reactive styles sensor data
  gets routed to each behaviour the needs that
  sensor, but is also available to the planner for
  construction of a task-oriented global world
  model.

38
Deliberation v.s. Planning

• Besides planning robot has to perform other
  tasks such as map making, performance
  monitoring, learning, etc.
• All these tasks together with planning are known
  as DELIBERATION

39
Sensing Organization
Deliberative functions Can eavesdrop Can have
their own Sensors Have output which Looks like
a sensor Output to a behavior (virtual sensor)
40
Hybrid Behaviors

• Behaviors are extended to
• Reflexive
• Innate
• Learned
• (Just like in ethology)

41
Architectures Common Functionality

• Mission planner
• Cartographer
• Sequencer agent
• Behavioral manager
• Performance monitor/problem solving agent (fairly
  rare)

42
Several Hybrid Approaches Have Been Developed

• AuRA (Arkin 1986)
• Atlantis (Gat 1991)
• Sensor-Fusion Effects (SFX) (Murphy 1996)
• 3-Tiered (3T) (JPL1990s)
• Saphira (Konolige 1998)
• Tack Control Architecture (Simmons 1997)
• Planner-Reactor (Lyons and Hendriks 1992)
• Procedural Reasoning System (PRS) (Georgeff and
  Lansky 1987)
• SSS (Connell 1992)
• Multi-Valued Logic (Saffiotti 1995)
• SOMASS Hybrid Assembly System (Malcom and
  Smithers 1990)
• Agent Architecture (Hayes-Roth 1993)
• Etc., Etc.

43
Georgia Tech TMR(Tactical Mobile Robot) Robots
44
NAVIGATION

• Topological Navigation Qualitative Navigation
• Metric Navigation Quantitative Navigation
• Navigation Algorithm usually is part of
  Deliberative part of Hybrid Architecture.

45
Qualitative Navigation uses Landmarks
46
floor plan
Gateway is an opportunity to change path heading
relational graph
Relational Methods Nodes landmarks,
gateways, goal locations Edges navigable path
47
Quantitative Navigation

• Want to get from one point to another with an
  optimization criteria
• Minimize Time
• Minimize Energy
• Minimize Distance
• Etc.

48
Space Representation Voronoi Graphs

• Imagine a fire starting at the boundaries,
  creating a line where they intersect,
  intersections of lines are nodes
• Result is a relational graph

49
Space Representation Rectangular Graph
50
A Algorithm
51
Discussion
Questions ?
52
References

• Murphy R.R. Introduction to AI Robotics, MIT
  Press, 2000
• http//www.policyalmanac.org/games/aStarTutorial.h
  tm (Accessed 2/6/06)
• http//robots.stanford.edu/ (Accessed (2/6/06)
• http//www.ghostriderrobot.com/index.php?idrobot
  (Accessed 2/6/06)