The failure of highway safety systems to perform in the real world as designed and tested under laboratory conditions has limited the effect of such systems in reducing the severity of highway collisions. In some instances, these safety systems have actually increased vehicle damage and occupant injuries.

This paper attempts to identify some of the missing design criteria which are leading to the inservice failure of laboratory-proven techniques. Examples of various appurtenance problems are given with explanations of why the systems did not perform as intended. The heart of the problem seems to lie at the design phase of planning a highway system and in maintaining the roadway and safety devices after construction. Only by investigating real-world accidents to understand better the effects of deviations from laboratory test conditions can useful information be collected to assist the designer as well as the appurtenance implementor and maintainer.