SOAR as a Cognitive Architecture for Modeling Driver

SOAR as a Cognitive Architecture for Modeling Driver Workload Randall Mauldin

Goal s To have onboard computer assistance that allows safe multi-tasking while driving. s Reduction of accidents and unsafe driving due to the distraction of secondary tasks proves to be a cause worth pursuing.

Introduction s Develop a Computational Cognitive Model of the driving task to allow a safer and more efficient driving experience.

How? s Develop a Cognitive Process Model (CPM) of a basic driver workload. s The CPM will take in to account various driver tasks and interpret their demand on cognition. s Develop computational specifications and implement them into a Cognitive Modeling Architecture.

Possibilities for a CPM

What is Driver Distraction? s Driver distraction lacks a precise, scientific definition. s Defined based upon four components: Impact, Agent, Mechanism, and Type.

Impact and Agent s “A driver is delayed in the recognition of information necessary to safely maintain the lateral and longitudinal control of the vehicle (the driving task)” s “Due to some event, activity, object or person, within or outside the vehicle”

Mechanism and Type s “That compels or tends to induce the driver’s shifting attention away from fundamental driving tasks” s “By compromising the driver’s auditory, biomechanical, cognitive or visual faculties, or combinations thereof”

SOAR s State Operator and Result s Created by John Laird, Allen Newell, and Paul Rosenbloom at Carnegie Mellon University in 1983. s The “state” is the situation that needs to be solved. s The “operator” is what changes the “state. ”

SOAR Soar’s 7 step decision cycle

SOAR Structural model of Soar’s operation

Key Features s Capable of representing large complex rule sets s Learns in a problem-solving context s New rules created for shorter sequences (“chunking”)