Skip to content

Instantly share code, notes, and snippets.

@pm5

pm5/FOSS4G_Day_1.md

Last active September 16, 2015 09:40
Show Gist options
  • Save pm5/f5a5eef2945eb11626e1 to your computer and use it in GitHub Desktop.
Save pm5/f5a5eef2945eb11626e1 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
FOSS4G_Day_1.md

FOSS4G

Keynote 1

  • Government should provide an enabling environment for FOSS4G to survive the vicious cycle.
    • Supporting community
    • Supporting FOSS4G R&D
    • Supporting FOSS4G application
    • Supporting FOSS4G education
  • NGII

Using QGIS Climate Change in Philipines

  • Rice consumption of Filipinos. 2 mil tons import, the largest rice importer in 2010.
  • IRRI, Internation Rice Research Institute. Gene bank of 112,000 types of rice, world biggest rice collection. Home to some most eminent rice scientists.
  • But 300,000 km2 land area are mountainous. Only 43,000 km2 are used for rice production.
  • Underutilized farm infra. Conversion to residential, commercial, and industrial land. Typhoons (big impact). Rapid population growth 2% per year.
  • "Food Staple Sufficiency" program of Department of Agriculture.
  • Subanen Tribe, indigenous to Zamboanga peninsula area. Region 9.
  • Areas growing Adlai. This presentation study region 11.
  • Adaptability trials. Stronger spicies of rice.
  • Our mission is to determine the suitable area for growing Adlai.
  • Site-selection analysis for Adlai.
  • Slope + Elevation with QGIS -> Soil Type -> Area Suitable for Adlai Cultivation. Three variables.
  • Took 3 months.
  • Data source: soild type depart argirculture (1950 data)
  • Farmare's feedback: research labs of Adlai has a lot of data. But no farmer knows abaout Adlai.
  • Plan: steping stone of using GIS and FOSS to promote any crops an your government. GIS software is expensive and only the government can buy them. Our budget is low and just a 1 year program. It's a good start.
  • Climate changes are not really considered in the result.

Some thoughts:

  • Problem-solving mindset.
  • Where to get the data? Philipine goverment sponsered?

Protecting tho Planet with PostGIS

Miguel Torres, UNEP-WCMC

  • UNEP-WCMC, 120 people working. We are in the biodiversity branch. I work in the info tech depart where we show our results in research on the web.
  • http://www.protectedplanet.net/ showing all protected area in the world on the web
  • You can just download the data. We try to get all information of PA.
  • GitHub https://github.com/unepwcmc/ProtectedPlanet
  • File geodatabase. 210,000 PA. 91% with boundaries. Remain points with area as attributes.
  • Monthly release of the World database on PA.
  • Contact country, update database, created annual reports, build website, create tasks to convert to PostGIs, automatic statistics (saves a lot of working's time).
  • For international orders there is no official dataset so I just use the Natural Earth dataset.

Some thoughts:

  • http://www.protectedplanet.net/terms cannot redistribute without prior written permission. What are the obstacles?
    • Some of the country don't want to have the data for open use. If it is to our options we would open them.
    • The data is sent from the countries so we are still dependent on them.
    • You can e-mail to request more usage requirements.

Making Sense of Sensor Data

Aurelia Moser, CartoDB, @auremoser

citizen science, social machines, and os maps

  • Sensors aggregate data human eyes can not see. DMZ, Sea weeds, etc.
  • auremoser's github and bl.ocks.org. Ushahidi. Brick to break Internet blockages in Africa. CartoDB.
  • http://landquest.internewskenya.org/ reporters
  • http://opennews.kzhu.io/map/disputes/
  • http://www.eurobirdportal.org/ebp/en/
  • How to build sensitive city maps
    • sensors - object whose purpose is to detect events or changes in its environment, and then provide a corresponding ouptput, a type of transducer.
      • inexpensive
      • buildable
    • pyramid of inverse utility: - personal/human < ambient/environmental < active/tactile < remote sensing +
    • "our space program is expensive slow and crappy anth that's why it's magic" -- Tom Sachs
    • http://datacanvas.org/sense-your-city/
    • swissnex-urbandata-challeng. "[Data Canvas] shows pepelo how difficult it is to deal with this data and make sense of it."
    • Ubidorts. freeboards. Temboo. OpenAQ. Envirofacts. EONet. Fitbits/Jawbone. Withings. Smart fridges.
    • torque.js
    • http://johnny-five.io/
    • http://nodebots.io/

Ontogoly

  • Some common challengs of GIS data and visualization.
  • A different approach. A method for modeing data as ontolgies that can help us turn messy data to useful information.
  • A crash course

    • Ontologies data, and user data. Two diff pipelines of data.

    • What are ontology data? A style of programming that allows you to make human-like inferences.

      • is a ; has a ; is a , etc.

    • Two central concepts. Universal, a collection of geo locatinos; GeoEntity, individual entity within a universal collection.

          USA [Country]
    +-> Colorado [State]
      +-> ...

    • Prupose is the provide a central geographic relaction. Well define spatial and non-spatial relationships. No dependency on geometries. (less GIS)

    • Ontologios in the OPEN: RunwaySDK, GeoDashboard.

    • What are user data? Data coming in from the user, CSV, etc. They can have relationships with ontologies.

    • Final Ontological Relationships: Data -> GeoEntity -> Universal

    • Why? Allows us to map data in a generic way. Everyone's data can be piped into it as long there is some ontologies.

      • No Geom? No problem.
      • What about geometries. Used to visualize GeoEntities, but optional.

Citizen Science and Smart Cities

Maria Arias de Reyna @geocat_bv

  • Cobweb https://cobwebproject.eu/ collect data from citizens through social network
  • Challenges
    • Adapted (metadata) catalog
    • Data analysis
    • QA
    • Data protection and security/privacy
    • Too complex UI
    • Semantics

Thoughts

  • roadkill.tw
  • GitHub where?

Usability Engineering for Successful Open Citizen Science

Jan Alexander Wirwahn @ifgi Westfalische Wilhelms-Universitat Munster

  • Work mainly about citizen science. Participatory sensing. Pupil teacher working with opensource hardware.
  • SenseBox, Toolkit for technical projects. How to publish the measurements? OpenSenseMap.
  • New project goals
    • Establish citizen drivent sensor network. DIY construction kits for sensor stations.
    • Nation-wide coverage of measurements. New standar for sharing open sensor data.
  • Opensource https://github.com/sensebox/OpenSenseMap
  • Test run in 2014 ~50 citizens with senseboxes. Try to use cheap compartments.
    • 8 million measurements collected in one year.
    • most stations were disconnected after a few weeks.
  • Error sources. We did a user study tested:
    • Hardware wiring
    • Software installation
    • Registration process
    • Documentation material
    • more user friendly SenseBox as a result
  • New design. More like a product:
    • Environmental DIY sensor station
    • Covering values of interest for citizens
    • Deployed once
    • Continous measurements
    • Data published as open data
    • New case, water-protection
  • Next steps
    • Start official measurement campaign
    • Extend documentation continuously
    • Open development for creative community
    • Find partners and promote

Thoughts

  • Does the new design decrease the educational value of the original SenseBox?
    • Different products and audiences. We have SenseBox for kids/edu, and for citizens.
  • Calibrated.
    • We make "calibrating session" to help people start using their boxes.

Space-time Mapping

CartoDB

Planet Labs

  • Dove orbit the earth in 90 minutes, capture scenes covering 30-300 km2.
  • Lots of imaginery. How do we let people find images they care about?
  • We build Scenes explorer. Now in the 3rd iteration.
  • Defining our target core user experience
    • Define an area of interest
    • Be able to view imagery in the area over time
    • Be able to filter images by metadata
    • Select and download imagery of interest
  • Demo time!
  • Which tools do we use?
    • The Planet API
    • React/Reflux
    • OpenLayers 3
    • CartoDB
    • and many others
  • Plant API
    • Scenes, Mosaics, Tiles, Workspace management, Notifications
    • JavaScript and Python libraries
  • Real-time filtering
    • Filtering is relatively slow.
    • We tried to move filtering to web worker. But trasferring data between threads is slow.
    • Then we use transferable objects. Geobuf for scenes.
    • Further optimization, only transfer deltas.
    • Able to handle ~20k scenes with good performance.
  • Further improvements
    • Geobuf support to the API.
    • Transfer more work to the worker.
    • Open source.

BoF Crowdsource

  • Open Data Kit
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment