1

Creating an Accurate Global Public Domain Roads Data Set

Global Roads Open Access Data Set (gROADS), a project of the ICSU CODATA Global Roads Data Development Task Group

By Alex de Sherbinin, CIESIN, Columbia University, Working Group Chair

Presented by Glenn Hyman, CIAT, Working Group Vice Chair

Presented at the 2011 Annual Meeting of the Association of American Geographers (AAG)

Outline

• Background and update on the CODATA-Roads initiative

• The Strategy for Data Development– Compile best available public domain data and cleaning it

– Approach mapping agencies

– Research or pilot data development efforts

• Strategy Moving Forward

2

3

gROADS Goal• To develop a global roads open access

data set (gROADS) that is:

1. globally consistent model (UNSDI-T v.2)

2. spatially accurate (~50m positional accuracy)

3. topologically integrated

4. focused on roads between settlements (not streets)

5. up-to-date and with the possibility of frequent updates

6. well documented

7. freely distributed (on attribution only basis)

VMAP0 – best available public domain data set

4

CODATA-Roads organizational milestones

• Workshop representing users and producers of roads data at CIESIN/Lamont Campus (October 2007)

• CODATA working group formed out of subset of participants, subsequently adding two members (Jan 2008)

• Becomes a Task Group of UN-GAID e-SDDC (UN Global Alliance on ICT for Development Open Access to and Application of Scientific Data in Developing Countries) (May 2008)

• Listed as a deliverable of the Group on Earth Observations (GEO) 2009-2011 Work Plan (October 2008)

• Listed as a milestone in the Interim United Nations Spatial Data Infrastructure Framework (November 2008)

• Received NASA-SERVIR Grant (March 2009)• Received Gates AGCommons Grant (April 2009)• Promoted to Task Group by CODATA (November 2010) – see member list

on next slide

5

CODATA Roads Task Group –March 2011

CODATA-Roads partners (1)1. Center for International Earth Science Information Network (CIESIN) (WG

co-chair A. de Sherbinin) – data cataloging and cleaning, coordination function for SERVIR project, AGCommons

2. Centro Internacional de Agricultura Tropical (CIAT) (WG member G. Hyman) – contributing Latin America roads data

3. Geographic Information Support Team, ITOS/UTA (WG member K. Payne) – providing a baseline for gROADS v.1

4. GISCorps – providing volunteer GIS staff

5. Global Road Inventory Project (GRIP), Netherlands Environmental Assessment Agency (PBL) – data and methods sharing

6. Global Spatial Data Infrastructure Association (GSDI) (WG member H. Onsrud) – endorsed the network, advice on licensing

7. Google Mapmaker (F. Lamy and J. Pfund) – provided gROADS Kenya data for comparison purposes, provided feedback on MM

6

CODATA-Roads partners (2)7. iMMAP (WG co-chair O. Cottray) – AGCommons coordinator

8. International Management and Mine Action Programs (IMMAP) (WG co-chair O. Cottray) – lead on AGCommons project

9. National institute for Advance Industrial Science and Technology (AIST), GEO Grid, and University of Tokyo (WG member K. Iwao) – collaborating on SERVIR project, developing semi-automated road feature extraction software and WFS-Transactional

10. OpenStreetMap (WG member M. Maron) – massive community of data developers committed to open data, gap filling in gROADS

11. Regional Center for Mapping of Resources for Development (RCMRD) (collaborating on SERVIR and AGCommons projects

12. Tracks4Africa (WG member J. Groenewald) – GPS tracks for spatial validation of existing data, advice on AGCommons work

7

CODATA-Roads partners (3)13. UNOSAT/SAFER project – developing RS-derived roads data for one or

more countries

14. World Food Programme – access to ITHACA, collaboration on AGCommons project

8

gROADS releases

Visit www.groads.org

CODATA Roads Data Catalog v.1 (Nov. 2009)

1. University of Georgia’s Information Technology Outreach Services (ITOS), compiled for the UN’s Geographic Information Support Team (GIST)

2. Netherland’s MNP Global Roads Inventory Project (GRIP)

3. Open Street Map

4. CIESIN data development activities

9

gROADS v.1 roads data set alpha release (November 2010)

with data from:

1. Catalog of 360+ national and regional data sets

gROADS v.1 final (summer 2011)

DEVELOPING DATA

10

Assembling Best Available Data

• Starting with ITOS data: A combination of VMAP1, DCW and national sources

• CIESIN is:– editing topology

– contributing new data by cleaning GPS derived data (e.g. Kenya), Landsat derived data (e.g. Nigeria), and Global Map (ISCGM) data

– combining data from multiple sources (e.g., DCW and Africover for Tanzania)

– conducting pilot projects that will provide new data for Ethiopia

11

Contacting national mapping agencies

Letter extract: “As a first step in creating this data set, we are contacting agencies such as your own to obtain permissions to incorporate existing national level data of suitable quality for inclusion in gROADS. In the next phase, we will be creating the infrastructure necessary to update these data in a distributed manner, so that those who contribute data can expect, within a three to four year time frame, to see improvements to their original dataset through the inputs of a wider community. Semi-automated means for facilitating distributed contribution of GPS tracks and digitizing tracks off of imagery are both being pursued as well as efficient means for gathering information about road characteristics. Since all data will be distributed free and clear of copyright restrictions, the data you contribute, plus any additions or enhancements, will be available to your organization at any time for extraction from gROADS. Any enhancements will be clearly documented in terms of the source of the updates. Thus, you obtain continuing visible credit for any contributions because each data source will be documented in the accompanying data documentation. Furthermore, your organization will be listed as a supporting data contributor on the gROADS wiki. You may highlight on your own website that your organization is a recognized official contributor of gROADS, and your organization will be granted authority to use the trademarked gROADScontributor logo in conjunction with this statement.

“We hope that you will agree that by contributing data to this initiative, you will not only benefit your own organization but also the wider community. If you are in agreement, we would be happy to send you a permissions form to be filled out, signed, and returned. We will be using the standardized Public Domain Dedication and License (PDDL) for this new and evolving dataset. Only in rare instances and for very large data collections will we accommodate other open access licensed contributions and these will be only on a temporary basis until such time as the data has been replaced through the distributed upgrading process.”

• An arrangement whereby governments give their roads data in the expectation of receiving better data in return

12

NASA-SERVIR Project (Ethiopia)

• Pilot tested a semi-automated road extraction tool developed by the Center for Spatial Information Science at University of Tokyo

• Target was ASTER imagery for Ethiopia

• The alpha version of the software was comparable to manual digitizing – line following algorithm underperforms due to similar

spectral signature for roads and surrounding land covers

• Average time per scene ~8hrs (depending on rural vs. urban and number of roads)

13

14

NASA-SERVIR project details

• Global Road Mapping Tool

• Data cleaning and development for East Africa

Google

Map

Aster

image

Aster

index

Seed points are inserted in:

1.ASTER image

2.Google map image

Similar to RoadTracker Commercial softwareSee: www.youtube.com/watch?v=azq0ZIIr6hI

AGCommons project (Ethiopia)• Output 1: A UNSDI-T compliant roads data set from GPS

tracks, along with agricultural features of interest

• Output 2: Software for a PDA tool that includes all fields of the UNSDI-T data model (based on Cybertracker)

• Project borrows approach of successful IMMAP projects in South Sudan and Liberia:

15

Incidental data collection: Engage third parties who are conducting missions to hard-to-reach places to collect data

Active data collection: Conduct field campaigns to collect data for regions that are missing

Incorporate data from third party sources where possible

RCMRD is collaborating on field campaigns and will disseminate the tool among member countries in East Africa

CIESIN is undertaking data cleaning and compilation

16

Customized Cybertracker Tool

Road Condition

Transport Point

Obstacle

Agricultural Point

WFP FDP

Strategy moving forward• Release gROADS v.1 with Web site including full documentation, FAQ, and

explanation of the CC0 license

• Solicit and evaluate inputs to a gROADS v.2 (2011-2012), and release gROADS v.2 (target: Winter 2013).

• Develop a non-commercial distribution global roads data set (gRDS) that includes data sets restricted to non commercial distribution (target: Spring 2012).

• Develop a gROADS Catalog v.2 which includes additional national and regional roads data sets (target: Fall 2012).

• Continue promoting and experimenting with road data development activities

• Continue interactions with OSM, GRIP, Google Mapmakers, GISCorps, World Resources Institute, and other groups interested in the development of improved roads data for applications and research.

17

Want to learn more or help out?

• Visit www.groads.org

• Sign up for the ROADSDATA discussion list to exchange information on data sets

• Send us your data!– Contact Alex de Sherbinin at amd155@columbia.edu

18

http://gismap.ciat.cgiar.org/egiron/GenerationAtlas/

We need GIS applications that don’t require users to be a GIS expert

http://africats.org/ http://agtrials.org/

Develop priority index for disease control by combining disease data and socio-economic data

Priority (D) =1/5 x A + 3/5 x B + 1/5 x C

A = BXW incidenceB = Bananas as % ofcaloric intakeC = % of food insecure people

Source: Bouwmeester, et al., 2008. ThePotential Benefits of GIS Techniques in Disease and Pest Control.

Develop priority index for disease control by combining the main two main cassava diseases

CMD

CBSD

Source: Legg and Bouwmeester, 2010. Cassava Disease Surveillance - Mapping Report Great Lakes Cassava Initiative.

Final comment

• Geospatial technologies should be built to leapfrog some traditional constraints (hardware, software, human resources, methods, data), in the same way that cell phones trumped land lines

– Cloud

– Applications that broadenuser base