A Productivity Model for Performance Evaluation of the UT Automated Road Maintenance Machine

Abstract

1: IntroductionIn recent years, several systems for automatically routing and sealing pavement surface cracks have been developed. Examples include (1) the CMU-UT Field Prototype (1992), (2) the CalDavis Field Prototype (1993), and (3) the UT Automated Road Maintenance Machine (ARMM) (1996). Through trial and error and over 8 years of perseverance , the UT ARMM has achieved an optimal balance between human and machine functions for automated pavement crack sealing[3 ]. For automation of pavement crack sealing, complete autonomy can be achieved , but at a cost and speed that is unacceptable [2]. This was apparent with the first CMU-UT field prototype, which used laser range sensing and machine vision to autonomously identify and map cracks.Currently, the ARMM combines machine vision and operator identification of the cracks to be sealed in order to map their exact locations in the machine's workspace . Recent field trials of the full scale crack sealer have indicated that automated pavement crack sealing is now technically, economically , and financially feasible. However, there is a need for periodic productivity studies of the full scale crack sealer to evaluate its performance as it is further refined.First, this paper briefly describes five specific functions for teleoperation of the ARMM. Then the paper will mainly focus on describing the methodology for determining the productivity of the ARMM. A mathematical model for measuring the system productivity has been developed based on the different types (longitudinal, trans-' Correspond to voungsuk(a