Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Automating Work Orders and Tracking Winter Snow Plows and Patrol Vehicles with Telematics Data

Anugunj Naman, Aaron Ault, Yaguang Zhang, James Krogmeier

TL;DR

The paper tackles the problem of labor-intensive and error-prone manual work-order tracking for winter road maintenance. It introduces an in-browser web application that ingests large-scale GPS telematics data, uses geohashing to map vehicle points to road segments, and computes per-segment work hours to automate work-order creation and verification. Key contributions include geohashing-based nearest-road identification, segment-level computed hours, validation workflows, and rich visualizations, all running client-side to preserve privacy. The approach promises reduced manual labor, improved accuracy, and better resource planning for INDOT and similar agencies, with future work focusing on non-overlapping segment representations and additional data sources to further refine driver-based work records.

Abstract

Winter road maintenance is a critical priority for the Indiana Department of Transportation, which manages an extensive fleet across thousands of lane miles. The current manual tracking of snowplow workloads is inefficient and prone to errors. To address these challenges, we developed an in-browser web application that automates the creation and verification of work orders using a large-scale GPS dataset from telematics systems. The application processes millions of GPS data points from hundreds of vehicles over winter, significantly reducing manual labor and minimizing errors. Key features include geohashing for efficient road segment identification, detailed segment-level work records, and robust visualization of vehicle movements, even on repeated routes. Our proposed solution has the potential to enhance the accuracy and granularity of work records, support more effective resource allocation, ensure timely compensation for drivers, alleviate administrative burdens, and allow managers to focus on strategic planning and real-time challenges. The web application can be accessed at https://github.com/oats-center/arrtrack/

Automating Work Orders and Tracking Winter Snow Plows and Patrol Vehicles with Telematics Data

TL;DR

The paper tackles the problem of labor-intensive and error-prone manual work-order tracking for winter road maintenance. It introduces an in-browser web application that ingests large-scale GPS telematics data, uses geohashing to map vehicle points to road segments, and computes per-segment work hours to automate work-order creation and verification. Key contributions include geohashing-based nearest-road identification, segment-level computed hours, validation workflows, and rich visualizations, all running client-side to preserve privacy. The approach promises reduced manual labor, improved accuracy, and better resource planning for INDOT and similar agencies, with future work focusing on non-overlapping segment representations and additional data sources to further refine driver-based work records.

Abstract

Winter road maintenance is a critical priority for the Indiana Department of Transportation, which manages an extensive fleet across thousands of lane miles. The current manual tracking of snowplow workloads is inefficient and prone to errors. To address these challenges, we developed an in-browser web application that automates the creation and verification of work orders using a large-scale GPS dataset from telematics systems. The application processes millions of GPS data points from hundreds of vehicles over winter, significantly reducing manual labor and minimizing errors. Key features include geohashing for efficient road segment identification, detailed segment-level work records, and robust visualization of vehicle movements, even on repeated routes. Our proposed solution has the potential to enhance the accuracy and granularity of work records, support more effective resource allocation, ensure timely compensation for drivers, alleviate administrative burdens, and allow managers to focus on strategic planning and real-time challenges. The web application can be accessed at https://github.com/oats-center/arrtrack/
Paper Structure (13 sections, 10 figures, 1 table, 2 algorithms)

This paper contains 13 sections, 10 figures, 1 table, 2 algorithms.

Table of Contents

  1. Introduction
  2. Proposed System
  3. Web Application
  4. Algorithm Formulation
  5. Geohashing for Nearest Road Identification
  6. Compute Seconds on Road Segment for Vehicle on Day
  7. Discussion
  8. Time Granularity in Work Orders
  9. Calculation of Work Hours for Drivers
  10. Visualization of Overlapping Roads When Tracking Vehicles
  11. Future Work
  12. Conclusion
  13. Acknowledgements

Figures (10)

  • Figure 1: Initial view of the web application.
  • Figure 2: Sample GPS data.
  • Figure 3: Sample work order records. Note that multiple entries for the same vehicle on the same day may appear, as highlighted by the red rectangle.
  • Figure 4: Sample verification report. The "Match" column, highlighted by the red rectangle, indicates how closely the computed time aligns with the reported time.
  • Figure 5: Work order creation interface.
  • ...and 5 more figures