Firesim provides immersive visualisation, natural interaction and rule based autonomy, it allows flexible scenario definition by a specially designed description language.
The systems includes a sound system providing spacial sound (explosion, fire noise, crying people, …) as well as speech synthesis and speech recognition for verbal interaction. A physics subsystem simulates an approximation for localized fires models (zone model).
Firesim provides a multi-user training environement.
The trainee may virtually walk through the scenary. Walking is simulated by “walk-sensors” attached to the boots. Walking is possible only in the direction of view.
A modified nozzle (C-size) is used for simulating direct fire-fighting. Parameters like pressure and foam-generation is controlled by an operator. Success of attack is controlled by an extinguishing model.
Fire (flames, smoke, fog)
- flames are modelled as multi-particle system, flame size proportional to current heat release rate
- smoke: 3d transparent bubbles
- fog: volumetric fogging (via GLSL shader), visibility depends on height of trainee (the lower the head the better the visibility will be)
fogging can be controlled by the operator manually
- sound of fire, loudness proportional to the distance to the fire
- fire fighting
- modified C-size nozzle, vent is connected to a switch (water off, spray or jet)
- position and direction is tracked by a Polhemus Patriot system, either direct (sensor attached to the nozzle) or indirect (sensor attached to a P5 data glove);
parable of ejection is computed with respect to angle of ejection, pressure, rate of flow, air resistance and type of ejection (jet/spray)
- extinguishing agents: wate, heavy and medium foam
- as light-foam and CAFM as well as carbondioxide require much more complex computation (compression!) they are not covered by this system
- simulation of extinction: absorbtion of energy by extinguishing agent, visual feed-back (size of flame, …)
- currently only sound events
- radiation source
- specification of radiation source by activity and position
- physics engine simulates Geiger-Mueller counter and computes overall activity as superpostion of all sources;
feed-back is given by sound (beeps) and digital HUD (head up display)
- (insured and immobile) persons
- immobile (insured) persons are show as texture
- depending on the grade of insury, sounds (crying, moaning, …) are produced (superposition!)
- mobile person (controlled by script-engine)
- natural talking (speech synthesizer (mbrola) and speech recognition (julius))
- hooter, sirenes, helicopter, …
- radio traffic
- director of operation (operator) and fire-fighter (trainee) may communicate using a head-set