A CAD‐based feature recognition system for escape route planning

The main objective of this paper is to introduce the development of a novel feature recognition system for escape route planning. The genetic algorithm (GA) can be a useful tool in solving route planning problems.

A computer‐aided design (CAD) system is proposed to help to predict the best escape route. The system is based on a feature recognition process, which will scan a CAD drawing to determine the centroid points of corridors, and determine the best escape route with the help of GA.

The feature recognition system provides a quick method of determining the available pathways for use in escape route planning.

Factors such as multiple story buildings, number of occupants, or fire spread are not considered. Thus the system will be useful in the initial stages of escape route planning.

The system is limited to perpendicular walls, which should be of the same thickness. Doorways should also be a constant size. Thus the system will be useful in determining where escape routes exist, and attempt to determine the best route based on that information.

This paper shows how legacy drawings from various CAD systems may be reused to plan escape routes, rather than creating new drawings for the same purpose. This is useful to those in escape route planning, since the redrawing of plans using custom software is inefficient.