The Consolidated Edison Company of New York, Inc. (Con Edison) has dev
eloped a computer program that predicts the reliability of auto-loops
and optimizes loop design. This paper describes the application of thi
s program to auto-loops in an urban area, the calculation of loop-spec
ific failure data using a Bayesian data update process and the develop
ment of strategies for enhancing loop reliability. Reliabilities predi
cted using loop-specific data are compared to those obtained using gen
eric data. The paper also demonstrates two alternative means of enhanc
ing auto-loop reliability: adding protective devices to a loop and rea
ligning its main run.