This research demonstrates a simple and economical way to assess road pavement conditions through any common smartphone. The phone was safely mounted inside a vehicle and sensor readings had been recorded during vehicle travel along different road segments. The vertical movement had been computed from the phone’s accelerometer and its location had been monitored through GPS. These two measurements made it feasible to predict pavement conditions without access to expensive instruments. As a data observation tool, a python script calculated Roughness Index (RI) based on vertical acceleration's root mean square (RMS) in short time windows. The output RI measures were categorized into three groups: smooth, moderate, and rough. The data were represented in green, yellow, or red dots onto a map and represented road quality well along the corridor. To better measure pavement condition, the campus road network was disaggregated into distinct sections based on natural corridor conditions such as intersections, curves, and direction changes. Breaking it down in this way enabled each section to be studied independently. The data showed strong variation between sections. Road segments near entrances and access points had repeatedly high Roughness Index (RI) values, indicating surface deterioration, patching, and unevenness. Straight internal segments and long corridors, on the other hand, had low RI values, which meant relatively smooth and well-maintenance pavement. This section-by-section analysis pinpointed areas in the campus network in need of maintenance and highlighted conditions based on pavement usage and location.
Imad ud din Ahmed (Wed,) studied this question.