A N A L Y T I C   V S .   G A M I N G S I M U L A T I O N S

Course Outline Challenger Disaster Disclaimer National Security Space Progs Mythical Man – Month Engineer – Mgr Conflicts Chief Scientist Complex Systems

EXAMPLE:   PROPOSED COLLISION AVOIDANCE SYSTEM FOR COMMERCIAL AIRLINERS •   Uses radar •   Detects other aircraft nearby     —   Are the aircraft too close?     —   Are they moving toward me?   •   Don’t move aircraft automatically       —   Warn pilot first       —   Warn air traffic control       —   Warn other pilot       —   Increase warnings as A/C get closer   •   Move aircraft automatically as last resort

O P E R A T O R   QUESTIONS

S Y S T E M  QUESTIONS

How much warning time do they need?     — Need reaction time     — Need decision time     — Need some for margin of error What procedures should they follow?     — Air traffic controller     — Pilot of 1st plane     — Pilot of oncoming plane How should they interact?     — Air traffic controller to 1st pilot     — Air traffic controller to 2nd pilot     — Pilot of oncoming plane

How accurate is the radar?   How soon can it detect oncoming A/C?       — How big do they have to be?       — How fast can they move?   How fast can planes move out of the way?       — Depends on size of plane       — Depends on engine capability   How close can planes come?       — Depends on size of plane       — Depends on response to turbulence

Answers to these questions generally depend on human factors, including the amount of training the operators have

Answers to these questions generally depend on physics, material properties, transmitter power, etc.

TOP

G A M I N G   SIMULATIONS

A N A L Y T I C SIMULATIONS

Run in, or near, real time     — Must do for training     — To test human response     — Seconds matter Speed outweighs accuracy     — Table look-ups     — Lower fidelity Used for prototyping     — Command centers     — Chain of command Same input does not yield same output Interface relatively important     — Colors     — Icons     — Developing I/F may be major purpose Runs interactively Multiple users     — Pilot 1     — Pilot 2     — Air Traffic Controller

Time usually not a concern     — Hours     — Days Accuracy outweighs speed     — Detailed algorithms     — High fidelity Determine functionality     — Sensor imaging limits     — Lift/drag ratio     — Maneuverability Same input yields same output Final answer important Interface relatively unimportant Can run in batch or background     — User can wait for answer     — Graphics done via post-processor Single user     — One set of inputs     — One set of outputs

TOP

P U R P O S E   O F   S I M U L A T I O N S

ANSWER QUESTIONS ABOUT SYSTEM W/O BUILDING IT

— Is it technically feasible?     — How fast can it respond?

Generally answered by analytic sim

— How will we use it?     — How fast can users respond?

Generally answered by gaming sim

— How much will it cost?

Answered by combination of sims   With additional analysis

— Impact of HW/SW failures

Answered by combination of sims   With additional analysis

TOP

S I M U L A T I O N   M I S U S E

Knowing how to use a system (via gaming simulation) doesn’t tell you if it’s feasible Knowing a system is feasible (via analytic simulation) doesn’t tell you how to use it Neither type of simulation automatically tells you cost Both kinds are needed to estimate effect of HW/SW failures Both kinds are needed to estimate O&M costs     — And they need to be designed to do this     — Most simulations are not     — How often will system be run?     — Cost to run once

DOD TENDS TO OVERUSE OR MISUSE GAMING SIMULATIONS

— Interest in human-in-control     — Military training for command, not analysis             •   Certainly true for higher ranks             •   Tendency to do what you know     — Gaming often used as substitute for analysis     — Analytic simulations may not be high-fidelity

TOP