+ All Categories
Home > Documents > Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A...

Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A...

Date post: 27-Mar-2015
Category:
Upload: julia-silva
View: 213 times
Download: 0 times
Share this document with a friend
Popular Tags:
39
Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks to D. Choi, T. Konik, N. Li, D. Shapiro, and D. Stracuzzi for their contributions. This talk reports research partly funded by grants from DARPA IPTO, which is not responsible for its contents.
Transcript
Page 1: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Pat Langley

School of Computing and InformaticsArizona State University

Tempe, Arizona USA

A Unified Cognitive Architecturefor Embodied Agents

Thanks to D. Choi, T. Konik, N. Li, D. Shapiro, and D. Stracuzzi for their contributions. This talk reports research partly funded by grants from DARPA IPTO, which is not responsible for its contents.

Page 2: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Cognitive Architectures

A cognitive architecture (Newell, 1990) is the infrastructure for an intelligent system that is constant across domains:

the memories that store domain-specific content

the system’s representation and organization of knowledge

the mechanisms that use this knowledge in performance

the processes that learn this knowledge from experience

An architecture typically comes with a programming language that eases construction of knowledge-based systems.

Research in this area incorporates many ideas from psychology about the nature of human thinking.

Page 3: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

The ICARUS Architecture

ICARUS (Langley, 2006) is a computational theory of the human cognitive architecture that posits:

It shares these assumptions with other cognitive architectures like Soar (Laird et al., 1987) and ACT-R (Anderson, 1993).

1. Short-term memories are distinct from long-term stores

2. Memories contain modular elements cast as symbolic structures

3. Long-term structures are accessed through pattern matching

4. Cognition occurs in retrieval/selection/action cycles

5. Learning involves monotonic addition of elements to memory

6. Learning is incremental and interleaved with performance

Page 4: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Distinctive Features of ICARUS

However, ICARUS also makes assumptions that distinguish it from these architectures:

Some of these tenets also appear in Bonasso et al.’s (2003) 3T, Freed’s (1998) APEX, and Sun et al.’s (2001) CLARION.

1. Cognition is grounded in perception and action

2. Categories and skills are separate cognitive entities

3. Short-term elements are instances of long-term structures

4. Inference and execution are more basic than problem solving

5. Skill/concept hierarchies are learned in a cumulative manner

Page 5: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Cascaded Integration in ICARUS

ICARUS adopts a cascaded approach to integration in which lower-level modules produce results for higher-level ones.

conceptual inference

skill execution

problem solving

learning

Like other unified cognitive architectures, ICARUS incorporates a number of distinct modules.

Page 6: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Goals for ICARUS

Our main objectives in developing ICARUS are to produce:

a computational theory of higher-level cognition in humans

that is qualitatively consistent with results from psychology

that exhibits as many distinct cognitive functions as possible

Although quantitative fits to specific results are desirable, they can distract from achieving broad theoretical coverage.

Page 7: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

An ICARUS Agent for Urban Driving

Consider driving a vehicle in a city, which requires:

selecting routes obeying traffic lights avoiding collisions being polite to others finding addresses staying in the lane parking safely stopping for pedestrians following other vehicles delivering packages

These tasks range from low-level execution to high-level reasoning.

QuickTime™ and aCinepak decompressor

are needed to see this picture.

Page 8: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

ICARUS Concepts for In-City Driving

((in-rightmost-lane ?self ?clane) :percepts ((self ?self) (segment ?seg)

(line ?clane segment ?seg)) :relations ((driving-well-in-segment ?self ?seg ?clane)

(last-lane ?clane) (not (lane-to-right ?clane ?anylane))))

((driving-well-in-segment ?self ?seg ?lane) :percepts ((self ?self) (segment ?seg) (line ?lane segment ?seg)) :relations ((in-segment ?self ?seg) (in-lane ?self ?lane)

(aligned-with-lane-in-segment ?self ?seg ?lane)(centered-in-lane ?self ?seg ?lane)(steering-wheel-straight ?self)))

((in-lane ?self ?lane) :percepts ((self ?self segment ?seg) (line ?lane segment ?seg dist ?dist)) :tests ( (> ?dist -10) (<= ?dist 0)))

Page 9: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Structure and Use of Conceptual Memory

ICARUS organizes conceptual memory in a hierarchical manner.

Conceptual inference occurs from the bottom up, starting from percepts to produce high-level beliefs about the current state.

Page 10: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Representing Short-Term Beliefs/Goals

(current-street me A) (current-segment me g550)(lane-to-right g599 g601) (first-lane g599)(last-lane g599) (last-lane g601)(at-speed-for-u-turn me) (slow-for-right-turn me)(steering-wheel-not-straight me) (centered-in-lane me g550 g599)(in-lane me g599) (in-segment me g550)(on-right-side-in-segment me) (intersection-behind g550 g522)(building-on-left g288) (building-on-left g425)(building-on-left g427) (building-on-left g429)(building-on-left g431) (building-on-left g433)(building-on-right g287) (building-on-right g279)(increasing-direction me) (buildings-on-right g287 g279)

Page 11: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

((in-rightmost-lane ?self ?line) :percepts ((self ?self) (line ?line)) :start ((last-lane ?line)) :subgoals ((driving-well-in-segment ?self ?seg ?line)))

((driving-well-in-segment ?self ?seg ?line) :percepts ((segment ?seg) (line ?line) (self ?self)) :start ((steering-wheel-straight ?self)) :subgoals ((in-segment ?self ?seg)

(centered-in-lane ?self ?seg ?line)(aligned-with-lane-in-segment ?self ?seg ?line)(steering-wheel-straight ?self)))

((in-segment ?self ?endsg) :percepts ((self ?self speed ?speed) (intersection ?int cross ?cross)

(segment ?endsg street ?cross angle ?angle)) :start ((in-intersection-for-right-turn ?self ?int)) :actions ((steer 1)))

ICARUS Skills for In-City Driving

Page 12: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

ICARUS Skills Build on Concepts

concepts

skills

Each concept is defined in terms of other concepts and/or percepts.

Each skill is defined in terms of other skills, concepts, and percepts.

ICARUS stores skills in a hierarchical manner that links to concepts.

Page 13: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Skill Execution in ICARUS

This process repeats on each cycle to give teleoreactive control (Nilsson, 1994) with a bias toward persistence of initiated skills.

Skill execution occurs from the top down, starting from goals to find applicable paths through the skill hierarchy.

Page 14: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Execution and Problem Solving in ICARUS

Executed plan

Problem

?

Skill Hierarchy

Primitive Skills

ReactiveExecution

impasse?

ProblemSolving

yes

no

Problem solving involves means-ends analysis that chains backward over skillsand concept definitions, executing skills whenever they become applicable.

Page 15: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

ICARUS Learns Skills from Problem Solving

Executed plan

Problem

?

Skill Hierarchy

Primitive Skills

ReactiveExecution

impasse?

ProblemSolving

yes

no

SkillLearning

Page 16: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Learning from Problem Solutions

operates whenever problem solving overcomes an impasse

incorporates only information available from the goal stack

generalizes beyond the specific objects concerned

depends on whether chaining involved skills or concepts

supports cumulative learning and within-problem transfer

ICARUS incorporates a mechanism for learning new skills that:

This skill creation process is fully interleaved with means-ends analysis and execution.

Learned skills carry out forward execution in the environment rather than backward chaining in the mind.

Page 17: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

ICARUS’ Memories and Processes

Long-TermLong-TermConceptualConceptual

MemoryMemory

Short-TermShort-TermBeliefBelief

MemoryMemory

Short-TermShort-TermGoal MemoryGoal Memory

ConceptualConceptualInferenceInference

SkillSkillExecutionExecution

PerceptionPerception

EnvironmentEnvironment

PerceptualPerceptualBufferBuffer

Problem SolvingProblem SolvingSkill LearningSkill Learning

MotorMotorBufferBuffer

Skill RetrievalSkill Retrievaland Selectionand Selection

Long-TermLong-TermSkill MemorySkill Memory

Page 18: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

An ICARUS Agent for Urban Combat

QuickTime™ and aCinepak decompressor

are needed to see this picture.

Page 19: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

ICARUS Summary

includes hierarchical memories for concepts and skills;

interleaves conceptual inference with reactive execution;

resorts to problem solving when it lacks routine skills;

learns such skills from successful resolution of impasses.

ICARUS is a unified theory of the cognitive architecture that:

We have developed ICARUS agents for a variety of simulated physical environments, including urban driving.

However, it has a number of limitations that we must address to improve its coverage of human intelligence.

Page 20: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Challenge 1: Arbitrary Behaviors

Retrieving relevant candidate skills for execution

Determining when skill execution should terminate

Constructing new skills from successful solutions

ICARUS indexes skills by the goals they achieve; this aids in:

But these goals can describe only instantaneous states of the environment, which limits ICARUS’ representational power.

For example, it cannot encode skills for complex dance steps that end where they start or the notion of a round trip.

Page 21: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Incorporating Temporal Constraints

Concepts that indicate temporal relations which must hold among their subconcepts

Skills that use these temporally-defined concepts as their goals and subgoals

A belief memory that includes episodic traces of when each concept instance began and ended

To support richer skills, we are extending ICARUS to include:

We are also augmenting its inference, execution, and learning modules to take advantage of these temporal structures.

Page 22: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Any round trip from A to B involves:

First being located at place A

Then being located at place B

Then being located at place A again

We can specify this concept in the new formalism as:

The Concept of a Round Trip

((round-trip ?self ?a ?b) :percepts ((self ?self) (location ?a) (location ?b)) :relations ((at ?self ?a) ?start1 ?end1

(at ?self ?b) ?start2 ?end2 (at ?self ?a) ?start3 ?end3)

:constraints ((≤ ?end1 ?start2) (≤ ?end2 ?start3)))

Page 23: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

The inference module automatically adds episodic traces like:

Episodes and Skills for Round Trips

((round-trip ?self ?a ?b) :percepts ((self ?self) (location ?a) (location ?b)) :start ((at ?self ?a) ?start1 ?end1) :subgoals ((at ?self ?b) ?start2 ?end2

(at ?self ?a) ?start3 ?end3) :constraints ((≤ ?end1 ?start2) (≤ ?end2 ?start3)))

The execution module compares these to extended skills like:

[(at me loc1) 307 398] [(home loc1) 200 …][(in-transit me loc1 loc2) 399 422] [(office loc2) 220 …][(at me loc2) 422 536] [(at me loc1) 558 …]

This checks their heads and uses constraints to order subgoals.

Page 24: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Challenge 2: Robust Learning

Solving novel problems through means-ends analysis

Analyzing the steps used to achieve each subgoal

Storing one skill clause for each solved subproblem

ICARUS currently acquires new hierarchical skill clauses by:

However, this mechanism has two important limitations:

It can create skills with overly general start conditions

It depends on a hand-crafted hierarchy of concepts

We hypothesize that a revised mechanism which also learns new concepts can address both of these problems.

Page 25: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Forming New Concepts

Create new conceptual predicates and associated definitions for start conditions and effects of acquired skills

That are functionally motivated but structurally defined

That extend the concept hierarchy to support future problem solving and skill learning

To support better skill learning, we are extending ICARUS to:

Learned concepts for skills’ preconditions serve as perceptual chunks which access responses that achieve the agent’s goals.

Page 26: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

(clear A)

(unstacked B A)

(unstackable B A)

(clear B) (hand-empty)(on B A)

(clear C)

(unstacked D C)

(unstackable D C)

DCBA

B

DCA

Learning Concepts in the Blocks World

When the problem solver achieves a goal, it learns both a new skill and two concepts, one for its preconditions and one for effects.

The system uses a mechanism similar to thatin composition (Neves & Anderson, 1981) to determine the conditions for each one.

ICARUS uses the same predicate in two clauses if the achieved goals are the same and if the initially true subconcepts are thesame (for concept chaining) or the utilized skills are the same (for skill chaining).

This produces disjunctive and recursive concepts.

Page 27: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

(clear A)

(unstacked B A)

(unstackable B A)

(clear B) (hand-empty)(on B A)

(clear C)

(unstacked D C)

(unstackable D C)

((scclear ?C) :percepts ((block ?C) (block ?D)) :relations ((unstackable ?D ?C)))

DCBA

B

DCA

Learning Concepts in the Blocks World

ICARUS solves novel problems in a top-down manner, using means-endsanalysis to chain backward from goals.

But it acquires concepts from the bottom up, just as it learns skills.

Here it defines the base case for the start concept associated with the skillfor making a block clear.

Page 28: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

(clear A)

(unstacked B A)

(unstackable B A)

(clear B) (hand-empty)(on B A)

(clear C)

(unstacked D C)

(unstackable D C)

((scclear ?B) :percepts ((block ?B) (block ?C)) :relations ((scunstackable ?C ?B)))

((scclear ?C) :percepts ((block ?C) (block ?D)) :relations ((unstackable ?D ?C)))

DCBA

B

DCA

Learning Concepts in the Blocks World

This process continues upward as thearchitecture achieves higher-level goals.

Here ICARUS defines the recursive case for the start concept associated with the skill for making a block clear.

Page 29: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

(clear A)

(unstacked B A)

(unstackable B A)

(clear B) (hand-empty)(on B A)

(clear C)

(unstacked D C)

(unstackable D C)

((scclear ?B) :percepts ((block ?B) (block ?C)) :relations ((scunstackable ?C ?B)))

((scunstackable ?B ?A) :percepts ((block ?B) (block ?A)) :relations ((on ?B ?A) (hand-empty) (scclear ?B)))

((scclear ?C) :percepts ((block ?C) (block ?D)) :relations ((unstackable ?D ?C)))

DCBA

B

DCA

Learning Concepts in the Blocks World

Skills acquired with these learned conceptsappear to be more accurate than those created with ICARUS’ old mechanism.

Page 30: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

(clear A)

(unstacked B A)

(unstackable B A)

(clear B) (hand-empty)(on B A)

(clear C)

(unstacked D C)

(unstackable D C)

((scclear ?B) :percepts ((block ?B) (block ?C)) :relations ((scunstackable ?C ?B)))

((scunstackable ?B ?A) :percepts ((block ?B) (block ?A)) :relations ((on ?B ?A) (hand-empty) (scclear ?B)))

((scclear ?A) :percepts ((block ?A) (block ?B)) :relations ((scunstackable ?B ?A)))

((scclear ?C) :percepts ((block ?C) (block ?D)) :relations ((unstackable ?D ?C)))

DCBA

B

DCA

Learning Concepts in the Blocks World

Page 31: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Benefits of Concept Learning (Free Cell)

Page 32: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Benefits of Concept Learning (Logistics)

Page 33: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Challenge 3: Reasoning about Others

The framework can deal with other independent agents, but only by viewing them as other objects in the environment.

ICARUS is designed to model intelligent behavior in embodied agents, but our work to date has treated them in isolation.

But people can reason more deeply about the goals and actions of others, then use their inferences to make decisions.

Adding this ability to ICARUS will require knowledge, but it may also demand extensions to the architecture.

Page 34: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

An Urban Driving Example

You are driving in a city behind another vehicle when a dog suddenly runs across the road ahead of it.

You do not want to hit the dog, but you are in no danger of that, yet you guess the other driver shares this goal.

You reason that, if you were in his situation, you would swerve or step on the brakes to avoid hitting the dog.

This leads you to predict that the other car may soon slow down very rapidly.

Since you have another goal – to avoid collisions – you slow down in case that event happens.

Page 35: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Social Cognition in ICARUS

Imagine itself in another agent’s physical/social situation; Infer the other agent’s goals either by default reasoning or

based on its behavior; Carry out mental simulation of the other agent’s plausible

actions and their effects on the world; Take high-probability trajectories into account in selecting

which actions to execute itself.

For ICARUS to handle social cognition of this sort, it must:

Each of these abilities require changes to the architecture of ICARUS, not just its knowledge base.

Page 36: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Architectural Extensions

Add abductive reasoning that makes plausible inferences about goals – via relational cascaded Bayesian classifier

Extend the problem solver to support forward-chaining search – via mental simulation using repeated lookahead

Revise skill execution to consider probability of future events – using the desirability of likely trajectories

In response, we are planning a number of changes to ICARUS:

These extensions will let ICARUS agents reason about other agents and use the results to influence its own behavior.

Page 37: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Automating Social Cognition

Treating goals achieved via anticipation as solved impasses;

Analyzing steps that led to this solution to learn new skills;

Using these skills to automate behavior when the agent finds itself in a similar situation.

Although humans can reason explicitly about other agents’ likely actions, they gradually compile responses and automate them.

The ICARUS skill learning module should achieve this effect by:

Over time, the agent will behave in socially relevant ways with no need for explicit reasoning or mental simulation.

Page 38: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

Concluding Remarks

Represent concepts and skills with temporal relations and use them to execute arbitrary behaviors;

Acquire new predicates that extend the concept hierarchy and enable better skill learning;

Reason about other agents’ situations and goals, predict their behavior, and select appropriate responses.

ICARUS is a unified theory of cognition that exhibits important human abilities but that also has limitations.

However, our recent work has extended the architecture to:

These extensions bring ICARUS a few steps closer to a broad-coverage theory of higher-level cognition.

Page 39: Pat Langley School of Computing and Informatics Arizona State University Tempe, Arizona USA A Unified Cognitive Architecture for Embodied Agents Thanks.

End of Presentation


Recommended