VEgAS (Virtual Egress and Analysis System)

(updated  31st August 2007)

Keywords: simulating people, simulating crowds, simulating crowd dynamics, workshops

"It is unrealistic to insist upon the highest service standards if the available physical space or financial resources are not sufficient to accommodate the design. A much better approach is to recognize the difficulties enforced by such restraints, and to seek alternative design solutions or operating procedures that will alleviate them. The one overriding consideration that precludes acceptance of economic or space restraints is pedestrian safety."  Fruin. Pedestrian Planning and design.

Note: The original VEgAS system was developed using Superscape - this platform is no longer available in the market - the following text refers to the principles applied in the VEgAS system.

Emergency exit routes are there to facilitate the rapid exit (egress) of the building occupants. However, until a real emergency occurs they may be untested. We can drill staff in evacuation procedures, we can also regularly test fire alarms but research has shown that the human reaction to the initial stages of an emergency egress is the dominant factor of a successful emergency evacuation. The question we addressed with VEgAS was "How can we use simulations and virtual reality to design the safer, quicker and more economic egress routes in building design in places of public assembly?"

Testing an egress strategy is a formidable problem. There are no prescribed set of blueprints for emergency planning as every event could be different. However, there are guidelines and codes of practice which should be applied to general cases. But, egress efficiency during an emergency cannot be fully tested until a 'real' crisis occurs. By then it may be too late. Computer simulations are a method of overcoming this problem and to this end a range of network flow, crowd and egress analysis systems are available on the market.

A building may be engineered to remove various egress problems using the approved guidelines and computer simulations. Engineers should employ a wide range of these techniques and tools for safety analysis, including the appropriate government guidelines and knowledge of past incidents. But, despite the detailed analysis of past emergencies (for example fires) it is impossible to predict the exact nature and extent of the next major emergency. If these events were predictable, we could all make a fortune as insurance consultants.

If we could somehow perceive the design in an emergency, visualise the problems and practice various egress strategies, we would be able to perform qualitative risk assessment. Unfortunately it is not feasible to physically test every building for all possible emergencies and the effective on egress. It is also unethical to expose people to real emergencies to analyse their behaviour for example, during a fire. How then can we be sure the egress design is operationally effective and safe?

Are people unpredictable?

The Human Factor has a major influence on the estimate of complete egress. As it is unwise to generalise on possible future human behaviour how can we reliably assess the risk involved in a mixed ability population?

A possible solution to these problems was the application of Myriad which has ergodic analysis of space into account during normal and emergency egress situations. The VEgAS system allowed the engineer to enter a computer simulation of his design, experience an event, use this insight to improve the design, and then test various contingency arrangements. It had behavioural characteristics at an individual entity level. However the analytical intractability of this system meant that dozens of scenarios required testing. Myriad solves all of these problems by assessing the spatial dynamics required for a successful evacuation. It is a planning and design tool and is being tested around the world on many different large scale events.

Using commercial Virtual Reality software models of crowds passing through doorways can be created.	This allows the researcher to test interactive rules and the behaviour of the individuals.
The advantage of a Virtual Reality interface is the ability to position cameras to allow the researcher a "pedestrian's eye view".	Models of stations and turnstiles/gate arrays can be created and moved using simple screen based interfaces.
Models of crowds move on a 2D plane......	....but with VR the researcher can add the extra dimensions of space....
...and track the entities movement through time.

Tracking entities through space has problems in that the set-up of individual entities, testing, validation and uncertainty that the human characteristics are correct inevitably lead to lack of confidence in models.

 "I had no idea that there was an army of Fire Protection Specialists and others who were more interested in results to the decimal point. (An 'unhuman' objective that is impossible to attain.)" Dr. J. Fruin, (personal correspondence May 2002).

Dr. G. Keith Still has a PhD in Crowd Dynamics (interdisciplinary mathematics) and has worked with psychologists, fire safety engineers, architects, designers, venue planners, safety officers, police officers, health and safety, building control and govt. departments. He developed a pedestrian simulation system during his PhD research. Crowd Dynamics Ltd produce a range of crowd simulation software and pedestrian analysis tools for computer simulation of places of public assembly.