Skip navigation
View Cart

Connected and Automated Vehicles

Cover of Implications of Pavement Markings for Machine Vision
Implications of Pavement Markings for Machine Vision
  • Publication no: AP-R633-20
  • ISBN: 978-1-922382-25-2
  • Published: 25 September 2020

This report examines how longitudinal pavement markings affect automated steering functions and whether changes to design and maintenance practices could improve the support lane markings provide for current advanced drivers assistance systems and future automated vehicles.

Vehicles with more automated features, including Lane Keeping Assist, are gradually being introduced into Australia and New Zealand. Lane Keeping Assist has been shown to have road safety benefits, especially on rural roads. Automated steering functions rely on machine-vision systems which use pavement markings and roadside features, such as kerbs and shoulders, to align the vehicle to the road.

The report documents the results of a literature review, stakeholder consultation, on-road and off-road evaluations and data analysis. It includes preliminary costings of key recommendations to allow agencies to understand the benefits and impacts of modifying line and lane marking practices.

  • Summary
  • 1. Introduction
    • 1.1 Project Purpose
    • 1.2 Project Objectives
    • 1.3 Project Methodology
    • 1.4 Project Scope
  • 2. Literature Review
    • 2.1 Objectives
    • 2.2 Stage One – Modern Machine-vision System Performance for Lane Guidance
      • 2.2.1 On-board Technology for Lane Guidance
      • 2.2.2 Machine Vision-enabled Lane Guidance
      • 2.2.3 How Various Manufacturers Approach Lane Guidance
      • 2.2.4 Pavement and Line-marking Standards on which Modern Machine-vision Systems are
      • 2.2.5 Limitations of Machine-vision Lane Guidance
      • 2.2.6 Similar Projects and Potential Lessons
      • 2.2.7 Key Factors
    • 2.3 Stage Two – Exploring Key Factors that Impact Machine-vision Performance
      • 2.3.1 Summary of Findings/Possible Best Practice from Literature Review
      • 2.3.2 Measurements and Evaluation of Daytime and Night-time Visibility
    • 2.4 Stage Three – Understanding the Current Standards Relevant to Australia and New Zealand
      • 2.4.1 Australian and New Zealand Standards
      • 2.4.2 Longitudinal Line Marking Gap Analysis
      • 2.4.3 Items to Discuss in Stakeholder Consultation
    • 2.5 Stage Four – Estimate the Road Infrastructure Whole-of-Life Costs to Support Automated Steering Functions
      • 2.5.1 Lifecycle
      • 2.5.2 Maintenance
      • 2.5.3 Unit Price of Pavement Marking Materials
      • 2.5.4 Estimated Road Asset Length
      • 2.5.5 Items to Discuss in Stakeholder Consultation
    • 2.6 Literature Review Summary
  • 3. Stakeholder Engagement
    • 3.1 Process
      • 3.1.1 Literature Review Findings Relevant to Stakeholders
      • 3.1.2 Identify Knowledge Gaps
      • 3.1.3 Identify Stakeholders
      • 3.1.4 Survey Design
      • 3.1.5 Inputs to Methodology
    • 3.2 Findings
      • 3.2.1 Road Agencies
      • 3.2.2 Vehicle Manufacturers
      • 3.2.3 Other Stakeholders
    • 3.3 Methodology Implications
  • 4. Test Scenario Identification
    • 4.1 Findings from Literature Review and Stakeholder Consultation
    • 4.2 Test Scope Discussion
    • 4.3 Test Scenario Development
    • 4.4 Off-road Trial Parameters
      • 4.4.1 Test Scenario A – Longitudinal Line Marking with High Qd Value
      • 4.4.2 Test Scenario B – Exit Ramp Configurations
      • 4.4.3 Test Scenario C – Longitudinal Line Marking with Varying Qd Value
      • 4.4.4 Test Scenario CW – Water-covered Dashed Longitudinal Line Marking
    • 4.5 On-road Trial Parameters
  • 5. Data Capture
    • 5.1 Vehicles and Test Equipment
    • 5.2 On-road Test Cases – Quantitative Data Analysis
      • 5.2.1 Data Analysis Procedure
      • 5.2.2 Data Processing
    • 5.3 Off-road Test Cases – Capture and Analysis
  • 6. Test Results
    • 6.1 On-road Test Case Findings
      • 6.1.1 Overall Performance
      • 6.1.2 Retroreflectivity
      • 6.1.3 Line Width
      • 6.1.4 Pavement Type
      • 6.1.5 Longitudinal Line-marking Configuration
      • 6.1.6 Wet Roads at Night-time
      • 6.1.7 Qualitative Data Analysis
      • 6.1.8 Key Factor Findings
    • 6.2 Off-road Test Case Findings
      • 6.2.1 Dashed Line Spacing and Line Width
      • 6.2.2 Exit Ramp Configuration
      • 6.2.3 Contrast Ratio of Qd
    • 6.3 Observations – Vehicle Response
    • 6.4 Test Result Summary
    • 6.5 Trial Limitations and Lessons Learnt
  • 7. Cost Assessment for Improving Line Marking
  • 8. Findings and Recommendations
    • 8.1 Findings
    • 8.2 Recommendations
  • 9. Conclusions
  • References
  • Appendix A Cost Benefit Analysis
    • A.1 Summary
    • A.2 Introduction
      • A.2.1 Purpose
      • A.2.2 Scope
    • A.3 Economic Appraisal Approach
      • A.3.1 Cost Estimate Methodology
      • A.3.2 Benefit Analysis Methodology
      • A.3.3 Summary of the Base ‘Do-Minimum’ Case and Project Scenario’s
      • A.3.4 Assumptions
    • A.4 Pavement Marking Requirements and Maintenance Cycles for Machine Vision Systems
      • A.4.1 Prioritisation of Non-urban Roads
      • A.4.2 Daytime and Night-time Pavement Markings Visibility
      • A.4.3 Line Width
      • A.4.4 Lane Width
      • A.4.5 Maintenance Cycle
    • A.5 Cost Estimate
      • A.5.1 Scenario Development
      • A.5.2 Building a Cost Model
      • A.5.3 Scale of Network by State
      • A.5.4 Scenario 1
      • A.5.5 Scenario 2
      • A.5.6 Cost Estimates within the Economic Analysis Framework
    • A.6 Benefit Analysis
      • A.6.1 Costs
      • A.6.2 Estimating the Benefits
      • A.6.3 Benefits not Quantified
    • A.7 Economic Appraisal Results
      • A.7.1 Cost Benefit Analysis
      • A.7.2 Sensitivity Tests
      • A.7.3 Conclusion
    • A.8 State by State Cost Estimate Information
      • A.8.1 Scenario 1 – Cost Estimates for Australian states and New Zealand, with Sensitivity Analysis
      • A.8.2 Scenario 2 – Cost Estimate for New Zealand, with Sensitivity Analysis
      • A.8.3 Percentage of Total Vehicles with Lane Support
Related publications
Webinar: Implications of Pavement Markings for Machine Vision
WEB-R633-20
Guidance and Readability Criteria for Traffic Sign Recognition Systems Reading Electronic Signs
AP-R627-20
Future Vehicles 2030
AP-R623-20
View all related publications
Latest Connected and Automated Vehicles News
17 Oct 2023
Recommended changes to physical infrastructure practices to improve connected and automated vehicle support

Austroads has published a report that identified physical infrastructure practices that transport agencies could change to improve connected and automated vehicle support. The report addresses recommended changes to various design, placement, and maintenance aspects for line and pavement marking, electronic and non-electronic signs, and traffic signals.
28 Jul 2023
Maximising value and improving the quality of automated vehicle trials

Austroads has published new guidance to aid with the evaluation and reporting of automated vehicle (AV) trials. The guidance provides organisations developing, approving or delivering AV trials with direction to improve the quality of trials and achieve greater harmonisation across jurisdictions.
View all news