introduction unit 1: introduction to autonomous robotics · comp 4766/6778 (mun) course...

7
Unit 1: Introduction to Autonomous Robotics Computer Science 4766/6778 Department of Computer Science Memorial University of Newfoundland January 12, 2009 COMP 4766/6778 (MUN) Course Introduction January 12, 2009 1 / 25 1 Introduction What is Autonomous Robotics What is this Course About? Relationship to Other Disciplines 2 Major Paradigms The Model-Based Paradigm Behaviour-Based Robotics Probabilistic Robotics COMP 4766/6778 (MUN) Course Introduction January 12, 2009 2 / 25 What is Autonomous Robotics Autonomous? Comes from the Greek for self-willed Something which is autonomous operates independently of external controls Robots? Comes from the Czech robotnik, meaning ‘workman’ From Karl Capek’s play “Rossum’s Universal Robots” “A machine used to perform jobs automatically, which is controlled by a computer” [Cambridge Dictionary, 2006] “Autonomous robots are intelligent machines capable of performing tasks in the world by themselves, without explicit human control over their movements.” [Bekey, 2005] COMP 4766/6778 (MUN) Course Introduction January 12, 2009 3 / 25 Which one of these robots is more autonomous? Autonomous robotics is distinct from industrial robotics which is concerned with the operation of robots in highly controlled environments COMP 4766/6778 (MUN) Course Introduction January 12, 2009 4 / 25

Upload: others

Post on 28-May-2020

2 views

Category:

Documents


0 download

TRANSCRIPT

Unit 1: Introduction to Autonomous Robotics

Computer Science 4766/6778

Department of Computer ScienceMemorial University of Newfoundland

January 12, 2009

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 1 / 25

1 IntroductionWhat is Autonomous RoboticsWhat is this Course About?Relationship to Other Disciplines

2 Major ParadigmsThe Model-Based ParadigmBehaviour-Based RoboticsProbabilistic Robotics

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 2 / 25

What is Autonomous Robotics

Autonomous?

Comes from the Greek for self-willed

Something which is autonomous operates independently of externalcontrols

Robots?Comes from the Czech robotnik, meaning ‘workman’

From Karl Capek’s play “Rossum’s Universal Robots”

“A machine used to perform jobs automatically, which is controlled bya computer” [Cambridge Dictionary, 2006]

“Autonomous robots are intelligent machines capable of performingtasks in the world by themselves, without explicit human control overtheir movements.” [Bekey, 2005]

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 3 / 25

Which one of these robots is more autonomous?

Autonomous robotics is distinct from industrial robotics which isconcerned with the operation of robots in highly controlled environments

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 4 / 25

Which one of these robots is more autonomous?

The Mars Pathfinder rover Sojourner (1997) was primarily manuallydriven; The Mars Exploration rovers Oppourtunity and Spirit (2003 -present) exhibit limited (but increasing) autonomy

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 5 / 25

What is this Course About?

This course provides an introduction to the computational aspects ofautonomous mobile robotics

We will not consider the following in any detail:

The construction of a robot’s body beyond the layout of its wheelsThe dynamics of robot motion

i.e. How forces on a robot’s body lead to velocities

We will focus on how to program a robot to...

Move in a particular direction (kinematics & control)Interpret sensor data and infer information about the worldMaintain an estimate of positionNavigate through both known and unknown environments

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 6 / 25

Relationship to Other Disciplines

Computer Science

Artificial Intelligence (AI)Computer Vision (CV)Computational GeometryAlgorithms

Computer and Electrical Engineering

Signal ProcessingControl Systems

Mechanical Engineering

Psychology, Neuroscience, Biology

Biological insights for roboticsRobotic instantiations of models

Autonomous Robotics (AR) is an inherently interdisciplinary field

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 7 / 25

AR as a Distinct Field of Study

AR is not just an application area for the preceding disciplines

By contrast to AI or CV, AR is distinguished by its focus onlarge-scale space [Dudek and Jenkin, 2000]

Robots must operate within environments which are larger than therobot’s immediate sensory horizonRequires:

Incremental acquisition of knowledgeRecognition of placesEstimation of positionReal-time response

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 8 / 25

Major Paradigms

A few major paradigms in AR have emergedModel-Based Paradigm

Build and maintain a model of the world and use it for planning

Behaviour-Based Robotics

Forget about modelling the world—simple behaviours can interactthrough the environment to yield complex emergent behaviours

Probabilistic Robotics

Assume that sensor data and robot actions are corrupted by noise;Represent the world and the robot’s place within it through probabilitydistributions

This list is not exhaustive

The paradigms listed above are also not mutually exclusive (numeroushybrid approaches exist)

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 9 / 25

The Model-Based Paradigm

The model-based paradigmbegan in the late 1960’s andwas heavily influenced bysymbolic approaches to AI

e.g. Nilsson and others atSRI developed “Shakey”

Shakey operated in anenvironment speciallymodified to assist its visionsystem

Its task was to pushparticular objects from oneone place to another

Based on STRIPSwww.ai.sri.com/shakey

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 10 / 25

STRIPS (STanford Research Institute Problem Solver)

Best illustrated using a “blocks world” environment[Luger and Stubblefield, 1998]

Environmental state described by a set of predicatesontable(a) on(b,a) clear(b)

ontable(c) on(e,d) clear(c)

ontable(d) gripping() clear(e)

Operations in the world represented by operations on these predicates:pickup(X), putdown(X), stack(X,Y), unstack(X,Y)

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 11 / 25

An operation such aspickup(X) affects the statedescription set as follows:

if gripping() ∧ clear(X) ∧ ontable(X)

add: gripping(X)

delete: ontable(X), gripping()

(This operation is for picking up objects lying

directly on the table)

The state space is searchedfor the goal state

STRIPS implements asearch through state space

to find a sequence ofoperations that wouldtransform the initial stateinto the goal state

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 12 / 25

Deficiencies

A number of deficiencies of the model-based paradigm have beenidentified

The symbol-grounding problem: “the symbols with which the systemreasons often have no physical correlation with reality” [Arkin, 1998]The modelling process is difficult

Sensor data is noisy and ambiguousUpdating the model is expensive and error-proneWorld / model deviations render plans useless

Many of these deficiencies remain in current work; some may beintractable

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 13 / 25

Behaviour-Based Robotics (BBR) [Arkin, 1998]

Term coined in the 1980’s but roots stretch back much further

Cybernetics: The science of control and communications in bothanimal and machine

Norbert Wiener utilized control systems theory to understand naturalbehaviour (1940’s)W. Grey Walter built a robotic tortoise exhibiting the followingbehaviours (1953)

Wander (lowest priority)Head toward a weak lightBack away from a bright lightAvoid obstacles (highest priority)

Robot acted on the highest priority applicable behaviourAbove the battery charger was affixed a strong light; when charge waslow this light was perceived as weakThus, a fully charged tortoise would back away from the bright chargerand begin to “explore” its world; When discharged it would return tothe apparently weak light of the charger

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 14 / 25

Braitenberg Vehicles [Arkin, 1998]

Valentino Braitenberg devised thought experiments to illustrate thatcomplex behaviour could result from very simple mechanisms (1984)

The Subsumption Architecture [Brooks, 1986]

Rodney Brooks proposed a behaviour-based approach called thesubsumption architecture [Brooks, 1991]

Brooks criticized the model-based functional decomposition

The tight coupling between layers leads to problems:

Errors made by earlier layers propagate to subsequent layersNo possibility for parallelismOverall update cycle is slowThe introduction of a new behaviour requires the modification of all

layers

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 17 / 25

The subsumption architecture organizes behaviours into vertical layerswith each layer acting out its own behaviour independently

There is no central controller; Each layer processes sensor data andcontrols actuators unless...

...suppressed or inhibited by another layerThus, there is a dynamic hierarchy of layers

New behaviours implemented as new layers without modifyingexisting layers (evolutionary growth)

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 18 / 25

“Intelligence without representation”: According to Brooks...

Intelligent behaviour emerges from a collection of simpler behaviours,appropriately interconnectedNo representation (i.e. model) is required: “use the world as its ownmodel”

Methodology:

Incrementally build in new behaviours—each capable of controlling therobot and achieving some taskRobots should be situated and embodied

Situated: Robot operates in the real world and is directly coupled to itthrough its sensors and actuatorsEmbodied: The robot’s brain should be housed within its bodySimulations allow experimenters to posit the same unrealisticassumptions made in an AI “blocks world”; A situated embodied robotcannot ‘fake it’

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 19 / 25

Deficiencies

Scalability

BBR may be suitable for low-level tasks but may not scale to moresophisticated tasksSome form of representation may be required for tasks where themoment-to-moment sensory information is insufficient

Thus, hybrid behaviour-based / model-based approaches are popular[Arkin, 1998]

Yet, neither approach addresses the pervasive influence of uncertainty

in robotics

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 20 / 25

Probabilistic Robotics [Thrun et al., 2005]

Increasingly popular since the mid-90’s; Roots of this paradigm canbe traced back to the invention of the Kalman filter (1960)

Premise:

Perception is uncertainThe results of robot actions are uncertainThese uncertainties should be represented explicitlyWe should represent “the world” as a probability distribution over allpossible worlds

“instead of relying on a single ‘best guess’ as to what might be thecase, probabilistic algorithms represent information by probabilitydistributions over a whole space of guesses” [Thrun et al., 2005]

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 21 / 25

An Example: Localization

Localization is the problem of estimating position w.r.t. the globalreference frame

In this example, a robot tries to localize itself within a 1Denvironment using a ‘door detector’ sensor and a map

Initially, the robot doesn’t know where it is, but does know itsorientation (facing to the right)

Notation:

x — the current position of the robotz — the current sensor observationbel(x) — robot’s belief (i.e. probability) that is at x , given both pastand current observations and movementsp(z |x) — probability of current observation given that robot is at x

Requires a map to know how likely an observation is at each location

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 22 / 25

Deficiencies?

Probabilistic robotics is the newest, most active paradigm in AR andis continuing to evolve at a fast pace; Thus, its success cannot yet befully characterized

Major challenge:

Navigation requires a mapThe representation of a probability distribution over all possible mapsrequires significant computational resources

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 24 / 25

References

Arkin, R. (1998).

Behavior-Based Robotics.MIT Press.

Bekey, G. (2005).

Autonomous Robots: From Biological Inspiration to Implementation and Control.MIT Press.

Brooks, R. (1986).

A robust layered control system for a mobile robot.IEEE Journal of Robotics and Automation, 2(1):14–23.

Brooks, R. (1991).

Intelligence without representation.Artificial Intelligence, 47:139–159.

Cambridge Dictionary (2006).

Cambridge online dictionaries.

Dudek, G. and Jenkin, M. (2000).

Computational Principles of Mobile Robotics.Cambridge University Press.

Luger, G. and Stubblefield, W. (1998).

Artificial Intelligence: Structures and Strategies for Complex Problem Solving.Addison Wesley.

Thrun, S., Burgard, W., and Fox, D. (2005).

Probabilistic Robotics.MIT Press.

COMP 4766/6778 (MUN) Course Introduction January 12, 2009 25 / 25