1 / 12

The Pothole Patrol

The Pothole Patrol. Using a Mobile Sensor Network for Road Surface Monitoring http://nms.csail.mit.edu/papers/p2-mobisys-2008.pdf. Timothy Werner. Problem . Potholes are everywhere Hard to find Car damage Lawsuits against the state Insurance claims. Solution. Pothole Patrol ( )

leann
Download Presentation

The Pothole Patrol

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. The Pothole Patrol Using a Mobile Sensor Network forRoad Surface Monitoring http://nms.csail.mit.edu/papers/p2-mobisys-2008.pdf Timothy Werner

  2. Problem • Potholes are everywhere • Hard to find • Car damage • Lawsuits against the state • Insurance claims

  3. Solution • Pothole Patrol () • Uses GPS and motion sensors • Deployable on existing vehicles • Taxis, garbage trucks, etc.

  4. Test Deployment • Attached to seven taxis in Boston for 10 days • Covered nearly 2500 distinct kilometers • Nearly 10000km total

  5. Test Deployment - Accelerometer Placement • Attached to car’s dashboard (glove compartment) • Accurate • Unobtrusive

  6. Test Deployment – Training Data • Trained for type of anomaly • Crosswalks and Expansion Joint • Railroad Crossing • Pothole • Manhole • Stopping • Turning

  7. Pothole Detection Algorithm • Speed • High-pass • z-peak • xz-ratio • Speed vs z ratio

  8. Pothole Detection Algorithm • Blacklisting • Speed bumps • Bridges • Etc • False Negatives – absence of reports does not mean a smooth road • Drivers naturally try to avoid pot holes • Roads are wide

  9. Evaluation • Tested on three data sets • Carefully labeled data • Loosely labeled data • The taxis driving around Boston

  10. Evaluation - Goals • Detect a low number of anomalies on smooth roads • Missing potholes is okay

  11. Evaluation • Classification accuracy on hand-labeled data • Performance improvement using loosely labeled training data • Performance on loosely labeled roads • Spot-checks on uncontrolled data

  12. Evaluation - Performance

More Related