Development of a smartphone application to measure pavement roughness and to identify surface irregularities

Islam, Md Shahidul

Permalink

https://hdl.handle.net/2142/89119 Copy

Description

Title

Development of a smartphone application to measure pavement roughness and to identify surface irregularities

Author(s)

Islam, Md Shahidul

Issue Date

2015-12-01

Director of Research (if dissertation) or Advisor (if thesis)

Buttlar, William G

Doctoral Committee Chair(s)

Buttlar, William G

Committee Member(s)

• Roesler, Jeffery R
• Work, Daniel B
• Vavrik, William R

Department of Study

Civil & Environmental Engineering

Discipline

Civil Engineering

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Pavement
• Roughness
• Smartphone Application

Abstract

Pavement roughness is an expression of the unevenness or disturbance in a pavement surface that adversely affects the ride quality of a vehicle. Roughness also affects user delay costs, fuel consumption, tire, and maintenance costs. Roughness is predominantly characterized in terms of International Roughness Index (IRI), which is often measured using inertial profilers. Inertial profilers are equipped with sensitive accelerometers, a height measuring laser, a distance measuring instrument, etc., to measure pavement profile. Modern smartphones are equipped with a number of sensors including a three-axis accelerometer, which has been utilized in this project to collect vehicle acceleration data using an android-based smartphone application. Two data analysis schemes have been developed to determine pavement profile from vehicle vertical acceleration data: a double integration and an inverse state space model. Acceleration data was double-integrated numerically to obtain a surrogate estimate of pavement profile based on the calculated vertical position of the vehicle cab. After noting a fairly significant underprediction of IRI for rough pavement sections with the double integration method, due in part to the dampening effects of the vehicle suspension, an inverse state space model was developed. This model enhances the double integration procedure by considering the physics of the mass-spring-damper system of the vehicle sprung mass as part of the back-estimation of road profile from vehicle cab acceleration. In addition, MATLAB and C# scripts were developed to estimate IRI from the pavement profile, using the procedure specified by ASTM. For initial validation, three test sites were selected to collect pavement profile using an inertial profiler along with acceleration collected using the smartphone application. These results demonstrated the potential for smartphone-measure IRI, as good correspondence to the inertial profiler was found for all but the roughest pavement investigated. The state space model was shown to provide significantly better estimates of IRI for rough pavement sections. Good repeatability between measurement replications was also noted, particularly when the space state model was used. For further validation, pavement roughness data was also collected using six smartphones and four vehicles. It was found that both the smartphone model and the vehicle used for data collection will affect the IRI measurement. However, averaged IRI values measured across all smartphones and vehicles were found to be in good agreement with the inertial profiler measured IRI for most of the pavement sections. The final phase of the study involved preliminary work in using the smartphone application for the purpose of pavement feature identification (bumps and potholes). Acceleration data collected using the smartphone application was filtered using an experimentally determined threshold value of 4 m/s2 to identify occurrences of significant localized distresses, and a MATLAB script has been developed to locate those distresses on a digital map. Finally, the smartphone application was used to collect roughness data over about 60 miles of roadway located in Champaign and Piatt County, IL, and measured IRI data has been integrated into a roadway network map using ArcGIS. In the roadway network map, every 0.1-mile pavement section has been highlighted with different colors based on measured roughness. It is hoped that the approach can be used to help reduce the cost of acquiring pavement roughness data for agencies and to reduce user costs for the traveling public by providing more robust feedback regarding route choice and its effect on estimated vehicle maintenance cost and fuel efficiency.

Graduation Semester

2015-12

Type of Resource

text

Permalink

http://hdl.handle.net/2142/89119

Copyright and License Information

Copyright 2015 Md Shahidul Islam

Owning Collections