A DECADE OF EXPERIENCE IN COMPUTER 
AIDED ROUTE SELECTION 
Dr. A. Keith Turner 
Environment Consultants, Inc. 
Lakewood, Colorado 
  
Introduction 
In 1966 Digital Terrain Models (DTM's) were of considerable research 
interest. Paul Roberts (8) had made some rather farsighted predictions 
in 1957 when he proposed the use of DTM's for highway location studies. 
By 1966 computers were being utilized in many areas of engineering de- 
sign research and early roadway design programs were available. Con- 
currently, new remote sensors were being developed and tested for many 
applications. Automated classification techniques were under development, 
partially in anticipation of the ERTS (LANDSAT I) satellite. 
At that time I decided to investigate methods of utilizing these new 
DTM databases for transportation planning. I assumed that the mechanics 
of producing them would be resolved and, since MIT was well into the ICES 
ROADS development, I elected to concentrate on the corridor selection 
phase. At that time corridor selection was mostly concerned with earth- 
work and other engineering costs, so there was a close tie to classic DTM 
concepts and less to the remote sensing aspects. 
Although not anticipated by me in 1966, passage of the National Environ- 
mental Policy Act (NEPA) in 1969 revolutionized the corridor selection 
process. Today the "best" design for a transportation system is no 
longer necessarily the one which produces the greatest reduction in 
travel time or the one which results in the lowest capital or user costs. 
Rather, it is the design that yields the highest social return on the 
transportation investment and reconciles most effectively the conflict- 
ing interests of the various groups affected by the proposal. 
Location engineers are faced with analyzing larger numbers of inter- 
acting and conflicting location factors. Decisions are required con- 
cerning the number of factors and the relative importance of all factors. 
Sensitivity analyses are required. The digital computer offers an 
efficient means of applying models which describe the regional environ- 
ment to the task of corridor selection. It should be noted that this 
requires the integration of a wide variety of digital models (or data- 
planes) each defining some component of the regional environment. 
The method I developed is defined as the Generalized Computer-Aïded 
Route Selection (GCARS) System. This paper traces the development of 
the GCARS System concepts from their first applications in 1968 to the 
present. As shown in Figure 1, the past decade can be divided into at