Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Introduction to Vector Tiles

Introduction to Vector Tiles

2022-11-25T15:00/16:00+07:00
FOSS4G Thailand 2022

Hidenori FUJIMURA

November 25, 2022
Tweet

More Decks by Hidenori FUJIMURA

Other Decks in Technology

Transcript

  1. Introduction to Vector Tiles
    UN Open GIS Initiative
    For FOSS4G Thailand 2022
    For JICA Knowledge Co-creation Program 2022
    Hidenori
    Map contents of this presentation are not necessarily authorized.
    Takeaway
    1

    View full-size slide

  2. Hello.
    • I am Hidenori, a government employee contributing to the UN through
    FOSS4G on web maps.
    • From OSGeo Japan chapter.
    • Lead, United Nations Vector Tile Toolkit (UNVT)
    • Lead, UN Open GIS Initiative Domain Working Group 7 (DWG 7)on Smart Maps
    • Working for the Geospatial Information Authority of Japan (GSI).
    • I am talking about Vector Tiles - an industry standard to distribute
    vector geospatial data on the web.
    • Presentation goal: to let us think together to keep web maps open - so
    that everyone can have access to geospatial information for smarter
    decision.
    2

    View full-size slide

  3. Presentation structure
    1. Examples
    2. Who we are
    3. UNVT Portable
    4. How we can combine efforts
    3
    to keep web maps open
    as UN Open GIS DWG 7

    View full-size slide

  4. 1. Examples
    1. Examples
    2. Who we are
    3. UNVT Portable
    4. How we can combine efforts
    4

    View full-size slide

  5. UN Maps in sync with OpenStreetMap 5
    Global coverage updated weekly.

    View full-size slide

  6. GSI Maps - size optimized and int’lized 6
    Open Source
    Government topographic maps in vector tiles

    View full-size slide

  7. GSI Maps - size optimized and int’lized 7
    Open Source
    Rich annotations

    View full-size slide

  8. GSI Maps - size optimized and int’lized 8
    Open Source
    Contour lines

    View full-size slide

  9. GSI Maps - size optimized and int’lized 9
    Open Source
    Detailed roads and tunnels

    View full-size slide

  10. GSI Maps - size optimized and int’lized 10
    Open Source
    Detailed in large scale

    View full-size slide

  11. GSI Maps - size optimized and int’lized 11
    Open Source
    1:2,500 roads and buildings

    View full-size slide

  12. GSI Maps - size optimized and int’lized 12
    Vector Tiles for the whole Japan area available up to 1:2,500.
    Open Source
    National dataset in around 30GB.

    View full-size slide

  13. optgeo/protomaps-vector-planet-odbl-port 13
    Experimenting PMTiles Open Source

    View full-size slide

  14. optgeo/etak - Estonian open geodata 14
    Open Source
    Demonstrating applicability in various data.

    View full-size slide

  15. optgeo/clear-optbv 15
    PoC on UN Clear Map + GSI Maps integration.
    Open Source
    Different sources together.

    View full-size slide

  16. optgeo/deck-optbv-intl - Deck.gl example 16
    Open Source
    Make sure our vector tiles work with different libraries.

    View full-size slide

  17. optgeo/jukyo-master - Gov’t address data 17
    Open Source
    Address data being developed by the Government of Japan

    View full-size slide

  18. optgeo/jukyo-master - Gov’t address data 18
    Open Source
    Use heat maps to show the data density.

    View full-size slide

  19. optgeo/jukyo-master - Gov’t address data 19
    Open Source
    Street level details.

    View full-size slide

  20. optgeo/jukyo-master - Gov’t address data 20
    Open Source
    House numbers when zoomed.

    View full-size slide

  21. 21
    Covering 3D - terrain and point clouds, too.

    View full-size slide

  22. optgeo/nagi-vi 22
    Buildings + terrain + orthophoto, with tilt-shift effect. Covering whole Japan.
    Open Source

    View full-size slide

  23. optgeo/togari (⿃狩) 23
    Light-weight visualization of point clouds. - ”Voxel Tiles”
    Open Source
    City area

    View full-size slide

  24. optgeo/togari (⿃狩) 24
    Light-weight visualization of point clouds. - ”Voxel Tiles”
    Open Source
    Mountains

    View full-size slide

  25. optgeo/tansei (淡⻘) COPC visualization 25
    Experimenting with open point cloud data from Virtual Shizuoka project.

    View full-size slide

  26. Click each box to see COPC data 27
    Open Source
    also to download them.

    View full-size slide

  27. COPC is easy. But it is big. 29

    View full-size slide

  28. COPC is easy. But it is big. 30
    300m x 400m
    Typical size = 100MB

    View full-size slide

  29. 2. Who we are
    1. Examples
    2. Who we are
    3. UNVT Portable
    4. How we can combine efforts
    31
    as DWG 7

    View full-size slide

  30. DWG 7
    Smart Maps
    1. We are a part of the UN Open GIS Initiative.
    2. We are Domain Working Group 7 (DWG 7) on Smart Maps.
    3. UN Vector Tile Toolkit is now a brand under DWG 7.
    32
    UN Open GIS Initiative DWG 7 on Smart Maps

    View full-size slide

  31. DWG 7
    Smart Maps
    UN Open GIS Initiative (http://unopengis.org/):
    1. The initiative to identify and develop an Open Source GIS bundle
    that meets the requirement of UN Operations.
    2. United effort to improve the effectiveness and efficiency of
    utilizing Open Source GIS around the world.
    33
    UN Open GIS Initiative DWG 7 on Smart Maps

    View full-size slide

  32. DWG 7
    Smart Maps
    Domain Working Group (DWG) 7 on Smart Maps (http://github.com/unopengis/7):
    ü We want to make it an open global practice community of engineers and
    operators who are keeping web maps open and diverse.
    ü Participants bring their high-level objectives with clearly defined owners and
    success criteria to get all the DWG members on the same page.
    ü We currently have 7 objectives.
    34
    UN Open GIS Initiative DWG 7 on Smart Maps

    View full-size slide

  33. DWG 7
    Smart Maps
    United Nations Vector Tile Toolkit (http://unvt.github.io):
    ü A collection of Open Source Software (OSS) to produce, host, style and optimize
    vector tiles for web mapping.
    35
    UN Open GIS Initiative DWG 7 on Smart Maps

    View full-size slide

  34. 2022 Upgrade from UNVT to Smart Maps
    • UNVT was a sub-project under UN Open GIS Initiative WG3 on Data
    Collection since 2018.
    • DWG7 on Smart Maps was newly established based on UNVT at the UN
    Open GIS Conference in Firenze, Italy 2022-08.
    • Smart Maps are user-oriented rather than technology-oriented.
    • Smart = Fast + Simple.
    • Smart Maps are not only about 2D vector data - also about 3D and real time.
    36

    View full-size slide

  35. Smart Maps - UNVT =
    DEM + COPC + LOD1
    1. Digital Elevation Model (DEM) in Terrain Tiles
    2. Point Cloud in Cloud Optimized Point Cloud (COPC)
    3. 3D City Models, firstly in LOD1, as Vector Tiles
    Smart Maps DWG takes advantage of base map expertise:
    Ø Data optimization based on the map scale.
    Ø Seamless integration of geospatial data.
    Vector Tiles continue to take the central part. But we also
    work with DEM, COPC, and LOD1 City Models.
    Metaphor
    37

    View full-size slide

  36. 7 Objectives under DWG7 38

    View full-size slide

  37. 7 Objectives under DWG7
    #1 UN operation by Taro U
    Basemap, terrain, satellite images
    with UN Global Service Centre.
    39

    View full-size slide

  38. 7 Objectives under DWG7
    #2 GSI operation by
    Hidenori
    Basemap, COPC, and
    capacity building.
    40

    View full-size slide

  39. 7 Objectives under DWG7
    #3 UNVT Portable by
    Taichi
    Capacity building and
    Disaster Response
    with Local Gov’ts
    41

    View full-size slide

  40. 7 Objectives under DWG7
    #4 Smart Campus by Santosh
    (Nascent, Ahmedabad, India)
    Smart Campus solution hosting
    42

    View full-size slide

  41. 7 Objectives under DWG7
    #5 3D Spatial ID by Takayuki
    (Geolonia, Japan)
    3D geospatial data free flow by
    extending vector tiles for 3D apps.
    43

    View full-size slide

  42. 7 Objectives under DWG7 44
    #6 3D Sensor lntegration by
    Hidenori
    Add Smart Maps values to existing
    sensor network infrastructure.

    View full-size slide

  43. 7 Objectives under DWG7 45
    #7 Core by Taro M
    Document, package, contribute to
    essential tools for Smart Maps.

    View full-size slide

  44. 7 Objectives under DWG7 46
    #7 Core by Taro M
    Document, package, contribute to
    essential tools for Smart Maps.
    YAML-based vector tile styling tool
    donated by Geolonia to UNVT.

    View full-size slide

  45. 7 Objectives under DWG7 47
    #7 Core by Taro M
    Document, package, contribute to
    essential tools for Smart Maps.

    View full-size slide

  46. Details: DWG7 Project Document
    Please see https://github.com/UNopenGIS/7/wiki/Project-Document
    48

    View full-size slide

  47. 3. UNVT Portable
    1. Examples
    2. Who we are
    3. UNVT Portable
    4. How we can combine efforts
    49

    View full-size slide

  48. UNVT Portable - Raspberry Pi map server 50
    Raspberry Pi - single-board PC
    developed by Raspberry Pi foundation.

    View full-size slide

  49. UNVT Portable - what and why
    • What
    • Single-board PC that works as a web map server.
    • Why
    • We needed an easy access to UNIX for capacity building.
    • We needed a cheap server that hosts data at the scale of 100GB.
    • GitHub Pages are good up to 1GB.
    • We use GitHub Pages for 100MB while we use UNVT Portable for 100GB.
    • We expect potential use for resource-limited environments.
    1. For disaster situations.
    2. For local governments.
    3. For capacity building.
    • I run a UNVT Portable server with 2TB of storage from home.
    51

    View full-size slide

  50. Benefit of UNVT Portable
    1. We can easily experiment
    technologies used in cloud-native
    environments.
    2. UNVT Portable is like a sky lantern
    because it can reach the “height of
    the cloud” in an affordable way.
    3. With multiple “sky lanterns” we can
    develop practical technologies
    combining existing tools.
    A) In-house object storage
    B) Battery-powered server
    Richy
    Metaphor
    52

    View full-size slide

  51. Experiment A: in-house object storage
    • COPC requires 100 times bigger storage.
    • Before COPC: 30GB - the size of microSD card.
    • After COPC: 3TB - the size of multiple external storage.
    • I am experimenting with MinIO - open source object storage software.
    53

    View full-size slide

  52. MinIO storing COPC 54
    3TB object storage with single-board PCs. Not very fast, but somewhat usable.
    And it works as a home server.

    View full-size slide

  53. Experiment B: off the grid
    • Can we run a UNVT Portable with mobile battery and solar panel?
    • Result is positive. Estimated running hours:
    55
    20100 mAh
    battery
    18 hours
    10000 mAh
    battery
    10 hours
    5000 mAh
    battery
    4 hours
    Tested with ANKER PowerCore.

    View full-size slide

  54. Application of UNVT Portable
    UNVT Portable = Cloud technology at hand
    • “Forward deployment” of the web map servers.
    üFor local governments and schools.
    üFor disaster situations.
    56
    Experiment
    Capacity Building
    Forward Deployment
    Application of UNVT Portable

    View full-size slide

  55. 4. How we can combine efforts
    1. Examples
    2. Who we are
    3. UNVT Portable
    4. How we can combine efforts
    57

    View full-size slide

  56. How we can combine efforts
    • DWG 7 is open.
    • It would be great if you participate!
    58
    Mission
    Community promises and values
    How to get involved
    UN Open GIS DWG 7

    View full-size slide

  57. Mission
    We build a global community of engineers
    and operators who are keeping web maps
    open and diverse.
    Current active participants include:
    United Nations Geospatial Information Section, United Nations Global
    Service Centre, Geospatial Information Authority of Japan, Aoyama
    Gakuin University, Crisis Mappers Japan, DRONEBIRD, Japan Flying Labs,
    Tokyo Cartographic, Nascent Infotechnologies, Geolonia.
    Participated by public, private, and academia.
    59

    View full-size slide

  58. Community Promises
    1. We treat people as equals. Everyone is welcome.
    2. We prefer open collaboration.
    3. Contributions are esteemed by their merit.
    4. We respect data producers.
    5. We criticize ideas, not people.
    6. We believe in fail forward to learn and improve.
    60

    View full-size slide

  59. Ground rules
    1. We move by empathy and trust, not by requests.
    2. It is OK to use local languages. It is nice to use common languages.
    61
    Some participants
    meets weekly using
    Japanese language.
    Why don’t you meet
    using Thai language?
    We can meet using
    English language, too.

    View full-size slide

  60. Community value statement
    1. We support participants to design, produce, style, host and optimize
    vector tiles.
    2. We support participants to cope with resource-limited environments.
    3. We support participants to learn about use cases.
    4. We support participants to have fun.
    5. We support participants to introduce new ideas.
    6. We support participants to help each other.
    7. We support participants to handle images, raster data (terrain tiles),
    point clouds, and 3D city models, too.
    62

    View full-size slide

  61. How to get involved
    Visit our repository to see what is happening.
    • https://github.com/unopengis/7
    Contact me. Ask any question.
    • Connect Hidenori Fujimura on Facebook.
    • Connect Hidenori Fujimura on LinkedIn.
    Bring your own objective and let’s collaborate.
    Thank you.
    63

    View full-size slide

  62. Introduction to Vector Tiles
    UN Open GIS Initiative
    For FOSS4G Thailand 2022
    For JICA Knowledge Co-creation Program 2022
    Hidenori
    Takeaway
    64

    View full-size slide