Keywords: crime prevention, CPTED, fuzzy set theory, parking garage, garage safety, artificial intelligence, fuzzy control, expert systems.

The authors developed a safety evaluation program for parking garages based on their investigation of the relation between environmental changes and people's perception of parking garage safety. The work is based on a study initiated by The Ohio State University in response to a campus parking garage crime rate that persisted at an unacceptably high level in spite of campus wide efforts to reduce crime. The authors developed a model based on crime statistics gathered by the OSU Police Department combined with results of an on-site survey evaluating Crime Prevention Through Environmental Design (CPTED) principles. The analysis included CPTED factors such as illumination, garage color, height of garage walls, the degree of transparency of elevators and stairways, and the exterior design of the campus parking garages. The evaluation showed that illumination was the most significant factors in users' perception of parking garage safety. As a result of this study, OSU implemented the recommended CPTED improvements. In the two years following the implementation of CPTED improvements, the average annual incidence of crime in the parking garage where the CPTED improvements had been made fell by more than half of the average annual incidence of crime in that same garage for the four years before the improvements were made.

Factors effecting perception of safety can be identified using CPTED principles. The perception of safety is represented by the safety index that is generated from a control mechanism that includes the fuzzification, inference engine, a rule base, and defuzzification of the inputs and outputs. Rule base and inference engine design was based on expert's knowledge or user's experience that inclusively includes environmental factors that may affect user's perception of safety. Results indicate that the level of user's perception of safety can be controlled at a desired level of user comfort if constraints governing the control mechanism were satisfied.

When assessing and evaluating parking garages, occupants employ experience- based judgment that although subjective and imprecise, is extremely useful. When designers employ the principles of CPTED, they also use subjective variables. An approach that captures the imprecise information in non-crisp expressions reflecting subjective and experience-based variables is based on the concept of fuzzy logic. This concept permits mathematical operations to be used to manipulate and operate on imprecise information or knowledge. Information on variables such as garage illumination and wall height was collected, fuzzified, and implemented into the development of many if-then rules forming the knowledge base of a model presented in this paper. These experience-based implication rules were developed and employed to evaluate the safety of the campus parking garage.

In this paper, the authors introduce an approach to control the safety level of a campus parking garage. This approach is different from the traditional Proportional- Integral-Derivative (PID) control because the input(s) and output(s) are not observable. The authors evaluate the perceived level of safety of the campus parking garage (safety index) as the input to the control mechanism that operates on CPTED variables and specifies changes in CPTED variables that will lead to the perception of safety desired by the user. This level of safety becomes the input to the control mechanism and the combinational effects of safety factors (change of physical characters) contribute to the output of the parking garage (performance).

Environment is a combination of physical characters that influence one's feeling under those circumstances. It is therefore important to examine the environment in order to facilitate the CPTED process. However, it is very difficult to change the existing physical characters of the parking garages in order to evaluate the outcome of the changes. Therefore, Virtual Reality (VR) was introduced to enable the re- examination of the level of safety of the targeted parking garage. A computer survey is taken when users examine the VR model. Data are then collected as input to the fuzzy control mechanism. This process repeats until the difference between the ideal and observed level of safety is within certain range. This controlling mechanism is achieved through the use of repetitive computer simulations.

Rule-based fuzzy controls have been implemented mainly in construction scheduling because of their capability of including uncertain factors in estimating the construction period. Such a fuzzy control system can be interpreted as a real- time expert system that combines qualitative linguistic symbols, their membership functions, and the experience-based if-then rules.

The authors found that perception of safety in campus parking garages is difficult to evaluate and control without the use of fuzzy logic and fuzzy control. Furthermore, the level of safety is linked to the funds available for improving safety. The level of safety can be controlled at a certain level when the funding is maintained on a certain level.

The safety evaluation program is a useful tool for planning and safety maintenance for the constructed facilities. Unlike many control programs that are based on physical changes to the targeted structure, this program attempts to explore the possibility of investigating the perception of safety that is invisible to the users and provide logical inference tool for evaluation of constructed facilities. The use of VR model not only provides a better understanding to those who have little experience in this field, but also functions as a good communication tool among the various users.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £82 +P&P)