口中 Information value Assessing the Benefit of Providing Information Average benefit from having perfect information (indicative of true state x), where one can select u to optimize for each x, is Vp=2i(xj)max; [V(xigui)I The best one can ever do from just knowing the a priori statistics V0=ma2px)W×u小 The difference(Vp-Vo) is the net benefit of having the inform- As a first approximation, the cost(in complexity of hardware/ software)of providing and/or of accessing the information (in or human time/ dis C=Eip(xi)c(xi) log211/p(xi)l Then should provide the information IFF(Vp-Vo)<C Information may become imperfect, e.g. state x changes with probability q to each other state before u takes effect. Then =2p×.1)max; [Vx q 2k#i [Vxk# Ui selected for i)) Then should provide information IFF(Vr-VO)<C (Courtesy of Thomas Sheridan. Used with permission
System Supervisory Control Computer Interface Display Control Sensors Direct Observation Information Value (Courtesy of Thomas Sheridan. Used with permission.)
小P.A T(x:r)
System Supervisory Control Computer Interface Display Control Sensors Direct Observation
口中 Information bandwidth · nformation rate 口Bts/Sec Information Density Raster Example 日(#Pⅸels)(#Bits/ Pixel)( Update Rate)
System Supervisory Control Computer Interface Display Control Sensors Direct Observation Information Bandwidth y Information Rate Bits/Sec y Information Density y Raster Example (# Pixels) ( # Bits/Pixel) (Update Rate)
口中 Task Performance&addh frame rate Depth Constant Task Performance Constant Bitrate Resolutio Diagram from Sheridan, Teleoperation
System Supervisory Control Computer Interface Display Control Sensors Direct Observation Task Performance & Bandwidth Frame Rate Constant Task Performance Color Depth Constant Bitrate Resolution Diagram from Sheridan, Teleoperation
Signal Detection Theol Originally Developed for Radar Threshold Detection Becomes the Basis for Alerting Theory · Signal versus Noise A-Scope Example
System Supervisory Control Computer Interface Display Control Sensors Direct Observation Signal Detection Theory y Originally Developed for Radar Threshold Detection y Becomes the Basis for Alerting Theory y Signal versus Noise y A-Scope Example