SlideShare a Scribd company logo

Geoverse Case Study: Using LiDAR to perform statewide inventory of sign assets

•Download as PPTX, PDF•
•
Merrick & Company
Merrick & Company

This document discusses a project by Greenman-Pedersen, Inc. (GPI) to use LiDAR technology to inventory all traffic signage along 14,000 lane miles of state-owned routes in Massachusetts. GPI proposed using high-resolution LiDAR and imagery collected from vehicles to extract sign data more efficiently than previous manual methods. GPI developed custom extraction software called HyRoad that integrates with the Geoverse point cloud visualization platform to allow navigating LiDAR data and extracting over 250,000 signs from over 50 terabytes of data. The project is currently over 40% complete.

Related slideshows

Digimap Update - Geoforum 2016 - Guy McGarva
Digimap Update - Geoforum 2016 - Guy McGarvaDigimap Update - Geoforum 2016 - Guy McGarva
Digimap Update - Geoforum 2016 - Guy McGarva
 
Geocap Water Column and Seafloor for ArcGIS - Oil and Gas seminar October 10th
Geocap Water Column and Seafloor for ArcGIS - Oil and Gas seminar October 10th Geocap Water Column and Seafloor for ArcGIS - Oil and Gas seminar October 10th
Geocap Water Column and Seafloor for ArcGIS - Oil and Gas seminar October 10th
 
Digimap Update - Guy McGarva
Digimap Update - Guy McGarvaDigimap Update - Guy McGarva
Digimap Update - Guy McGarva
 
1 of 18
Presented to: The International LiDAR Mapping Forum Geoverse Case Study: Using LiDAR to perform statewide inventory of sign assets February 24, 2015
About Greenman-Pedersen, Inc (GPI) • Multi-discipline engineering firm • Established in 1966 • 1100 employees • 20 offices • Transportation Asset Management
Project Overview • Traffic Sign Asset Management Project – Massachusetts DOT – Inventory of all signage on state-owned routes (14,000 lane miles) – Night-time Retro-reflectivity condition – Implement Asset Management System – Tie into existing Maximo system – 2 Year Project
GPI Proposal • GPI has completed over 50 similar projects • Proposal: Use high-res LiDAR / High res Inventory to conduct inventory • Value Added Proposition. Drive it once 1. Additional Assets 2. Visualization 3. Survey Grade Areas
Data Collection NIKON D800 • 36 MP Highest Resolution ML Calibrated Video/Still • Mapping Camera RIEGL VMX450 • Up to 1,100,000 measurements per second • Dual 360 FOV Scanners • Up to 5000 points/SM NIKON D800 RIEGL VMX450
Project Overview VueWorks Lidar files Photos Amazon EC2 Cloud (Virginia) •Drive Limits •Route Priority •High Resolution Areas •Drive Schedule Drive Plan •LiDAR /Camera Array •Data Delivery to GPI Data Collection •HyRoad •Daytime Condition •Load to VUEWorks Extraction •Comparison Panels •Nightrider Retroreflectivity Maximo Export GIS Road Construction Project Schedule MassDOT priority
Data Details • Inventory approximately 250,000 signs • Dataset is over 50 TB • 40 TB LiDAR • How to deal with it?
The Old Way • Data Extractors navigate through imagery to find assets. • Work in small sections • Cut cross sections • Perform measurements / assessment • Time consuming / Unorganized
Wouldn’t it be nice…. • Navigate seamlessly through the point cloud • View imagery and LiDAR in concert • Speed up extraction • ILMF last year
What Geoverse Did (and Didn’t) • Pros − Could handle large amounts of LiDAR data − Measuring Capability • Cons − Had no database tie in − Software Development Kit (SDK) $$$$
Solution: Geoverse API • Worked with Euclidean and Merrick to develop a Geoverse Programming Interface • Allow data to be passed between GPI developed extraction software and Geoverse • 2 key components − Control Geoverse point of view − Return measurements • Developed Custom Extraction software: HyRoad
HyRoad Software Demonstration
Current Project Status • Extracted over 70,000 signs • Night Assessed Over 50,000 signs • Driven over 40% of project roads • Planning spring drive missions if the snow ever melts….
Realized Benefits • Ease of locating signs in point cloud • Increased extraction productivity • Can view entire dataset in a single project • Compression • Opened several possibilities for data visualization
Geoverse Details • Data displayed using a 3cm pixel • Compression is approximately 20:1 • Each LAS file is converted to UDS format • 1 UDS / LAS per Route Direction
Lessons Learned • Compression allows for new horizons • Geoverse is not LiDAR • Processing…. • Procedures need to be altered • Work with good partners
MassDOT Geoverse Demo 688,978,485,040,032 points so far…
Questions Mark Day Director of Application Development Greenman-Pedersen, Inc. Engineering and Construction Services 325 West Main Street, Babylon, NY 11702 d 631.539.3705 mday@gpinet.com

More Related Content

Geoverse Case Study: Using LiDAR to perform statewide inventory of sign assets

  • 1. Presented to: The International LiDAR Mapping Forum Geoverse Case Study: Using LiDAR to perform statewide inventory of sign assets February 24, 2015
  • 2. About Greenman-Pedersen, Inc (GPI) • Multi-discipline engineering firm • Established in 1966 • 1100 employees • 20 offices • Transportation Asset Management
  • 3. Project Overview • Traffic Sign Asset Management Project – Massachusetts DOT – Inventory of all signage on state-owned routes (14,000 lane miles) – Night-time Retro-reflectivity condition – Implement Asset Management System – Tie into existing Maximo system – 2 Year Project
  • 4. GPI Proposal • GPI has completed over 50 similar projects • Proposal: Use high-res LiDAR / High res Inventory to conduct inventory • Value Added Proposition. Drive it once 1. Additional Assets 2. Visualization 3. Survey Grade Areas
  • 5. Data Collection NIKON D800 • 36 MP Highest Resolution ML Calibrated Video/Still • Mapping Camera RIEGL VMX450 • Up to 1,100,000 measurements per second • Dual 360 FOV Scanners • Up to 5000 points/SM NIKON D800 RIEGL VMX450
  • 6. Project Overview VueWorks Lidar files Photos Amazon EC2 Cloud (Virginia) •Drive Limits •Route Priority •High Resolution Areas •Drive Schedule Drive Plan •LiDAR /Camera Array •Data Delivery to GPI Data Collection •HyRoad •Daytime Condition •Load to VUEWorks Extraction •Comparison Panels •Nightrider Retroreflectivity Maximo Export GIS Road Construction Project Schedule MassDOT priority
  • 7. Data Details • Inventory approximately 250,000 signs • Dataset is over 50 TB • 40 TB LiDAR • How to deal with it?
  • 8. The Old Way • Data Extractors navigate through imagery to find assets. • Work in small sections • Cut cross sections • Perform measurements / assessment • Time consuming / Unorganized
  • 9. Wouldn’t it be nice…. • Navigate seamlessly through the point cloud • View imagery and LiDAR in concert • Speed up extraction • ILMF last year
  • 10. What Geoverse Did (and Didn’t) • Pros − Could handle large amounts of LiDAR data − Measuring Capability • Cons − Had no database tie in − Software Development Kit (SDK) $$$$
  • 11. Solution: Geoverse API • Worked with Euclidean and Merrick to develop a Geoverse Programming Interface • Allow data to be passed between GPI developed extraction software and Geoverse • 2 key components − Control Geoverse point of view − Return measurements • Developed Custom Extraction software: HyRoad
  • 13. Current Project Status • Extracted over 70,000 signs • Night Assessed Over 50,000 signs • Driven over 40% of project roads • Planning spring drive missions if the snow ever melts….
  • 14. Realized Benefits • Ease of locating signs in point cloud • Increased extraction productivity • Can view entire dataset in a single project • Compression • Opened several possibilities for data visualization
  • 15. Geoverse Details • Data displayed using a 3cm pixel • Compression is approximately 20:1 • Each LAS file is converted to UDS format • 1 UDS / LAS per Route Direction
  • 16. Lessons Learned • Compression allows for new horizons • Geoverse is not LiDAR • Processing…. • Procedures need to be altered • Work with good partners
  • 18. Questions Mark Day Director of Application Development Greenman-Pedersen, Inc. Engineering and Construction Services 325 West Main Street, Babylon, NY 11702 d 631.539.3705 mday@gpinet.com