WEBVTT

1
00:00:00.000 --> 00:00:05.000
It looks no different from a typical map site.

1
00:00:05.000 --> 00:00:09.000
The layers are listed in this list box.

1
00:00:09.000 --> 00:00:18.000
Each layer is a hyperlink, and each one refers to content from a different domain.

1
00:00:18.000 --> 00:00:26.000
For example, this is rainfall information from the Meteorological Office. As you can see, there are many layers.

1
00:00:26.000 --> 00:00:31.000
This is the original page being referred to.

1
00:00:31.000 --> 00:00:43.000
This is a weather chart. The weather chart is drawn in Stereographic Projection, but its transformation is also made in this way.

1
00:00:43.000 --> 00:00:50.000
This is a hazard map showing tsunami hazard levels from the Ministry of Land, Infrastructure, Transport and Tourism.

1
00:00:50.000 --> 00:00:52.000
This is the original page being referred to.

1
00:00:54.000 --> 00:01:07.000
This is information from live cameras on rivers. You can get not only the location of the cameras, but also their respective live images.

1
00:01:11.000 --> 00:01:17.000
Next, let's look at one of the effective use cases for hyperlayering.

1
00:01:17.000 --> 00:01:28.000
Let us superimpose a page that provides information on the damage caused by the Noto Peninsula earthquake of January 2024, using a synthetic aperture radar satellite.

1
00:01:28.000 --> 00:01:36.000
First, we trace the damaged area from this photo and overlay it on the map.

1
00:01:36.000 --> 00:01:42.000
When zoomed in, we can get a detailed damage for each house.

1
00:01:42.000 --> 00:01:46.000
However, this is difficult to do with the original page alone.

1
00:01:46.000 --> 00:01:55.000
As you can see, since hyperlayering can be expected to have great benefits, it may require a newly discussion on copyrights.

1
00:01:56.000 --> 00:02:03.000
In fact, I have built and tested HLA layers for more than five hundred different systems.

1
00:02:03.000 --> 00:02:10.000
They use different data and mapping technologies, so they cannot simply be plugged into the same map.

1
00:02:10.000 --> 00:02:15.000
To overcome these differences, HLA uses an architecture called Layers as Web Apps.

1
00:02:15.000 --> 00:02:21.000
Each layer is an independent Web App that can handle its own data and behavior.

1
00:02:21.000 --> 00:02:25.000
I kept building these small Web Apps, one by one.

1
00:02:25.000 --> 00:02:32.000
They can absorb the differences between the underlying systems, while working together on the same map.

1
00:02:32.000 --> 00:02:38.000
As a result, more than two hundred of these Layers as Web Apps are available for public release.

1
00:02:41.000 --> 00:02:45.000
Now, let me show you one more example from Canada.

1
00:02:45.000 --> 00:02:50.000
This content was built by the community, showing that Hyper Layering and Layers as Web Apps are

1
00:02:50.000 --> 00:02:54.000
now being explored beyond Japan.

1
00:02:54.000 --> 00:02:58.000
First, let me display it in the familiar Plate Carrée projection.

1
00:02:58.000 --> 00:03:02.000
But actually, we can do something more.

1
00:03:07.000 --> 00:03:13.000
This same content can also be displayed in the Lambert projection used for Canada.

1
00:03:13.000 --> 00:03:16.000
And this is the interesting part.

1
00:03:16.000 --> 00:03:23.000
A map projection can be implemented as a function inside a Layers as Web Apps layer.

1
00:03:23.000 --> 00:03:31.000
This means that layers using different projections can be integrated and displayed together in any projection.

1
00:03:33.000 --> 00:03:35.000
Now, let's look at another kind of layer.

1
00:03:35.000 --> 00:03:40.000
There is a powerful technique called QuadTree Composite Tiling.

1
00:03:40.000 --> 00:03:42.000
Let me show you what it can do.

1
00:03:42.000 --> 00:03:48.000
This time, I am using OpenStreetMap as a source of thematic data.

1
00:03:48.000 --> 00:03:56.000
OpenStreetMap is not only a great basemap. It is also a remarkably rich source of thematic data.

1
00:03:56.000 --> 00:04:01.000
Here are parking facilities. Even at this scale, we can see their spatial distribution.

1
00:04:01.000 --> 00:04:09.000
As we zoom in, individual features appear as vector points.

1
00:04:13.000 --> 00:04:22.000
And here are drugstores. Again, we can see their distribution first, and then the individual features.

1
00:04:22.000 --> 00:04:28.000
QuadTree Composite Tiling is particularly well suited to this kind of thematic data.

1
00:04:28.000 --> 00:04:35.000
When the data itself becomes the subject of the map, its spatial distribution becomes meaningful.

1
00:04:36.000 --> 00:04:39.000
So, how is this different from clustering?

1
00:04:39.000 --> 00:04:42.000
Here is the same data with clustering.

1
00:04:42.000 --> 00:04:47.000
Clustering is useful for finding individual points.

1
00:04:50.000 --> 00:04:54.000
Now, the same data with QuadTree Composite Tiling.

1
00:04:54.000 --> 00:04:59.000
Instead of simply grouping the points, it represents their spatial distribution.

1
00:04:59.000 --> 00:05:04.000
And we can also switch to a heatmap representation.

1
00:05:11.000 --> 00:05:16.000
And QuadTree Composite Tiling is not limited to point data.

1
00:05:16.000 --> 00:05:19.000
Here are bus routes as the subject of the map.

1
00:05:19.000 --> 00:05:23.000
The same approach works for linear features as well.

1
00:05:23.000 --> 00:05:27.000
There is another interesting aspect of QuadTree Composite Tiling.

1
00:05:27.000 --> 00:05:31.000
It is not tied to a particular implementation.

1
00:05:31.000 --> 00:05:36.000
It can be implemented as Layers as Web Apps, a static hyperdocument,

1
00:05:36.000 --> 00:05:39.000
or a dynamically generated service.

1
00:05:39.000 --> 00:05:45.000
The important point is that a layer is free to choose its own implementation.

1
00:05:48.000 --> 00:05:54.000
Moreover, the capabilities of Hyper Layering go beyond simply overlaying map layers.

1
00:05:54.000 --> 00:06:01.000
It even allows geospatial operations to be performed between arbitrary layers that are superimposed.

1
00:06:01.000 --> 00:06:04.000
First, let's display the layers of the clinics.

1
00:06:04.000 --> 00:06:13.000
Next, a layer of seismic intensity distribution for the Noto Peninsula earthquake of January 2024 is displayed.

1
00:06:17.000 --> 00:06:22.000
Then, we can perform operations between these layers to

1
00:06:22.000 --> 00:06:28.000
extract medical facilities that have been subjected to major quakes of intensity 6 or higher.

1
00:06:29.000 --> 00:06:34.000
The operations are performed on the client side with crawling.

1
00:06:34.000 --> 00:06:41.000
Completed. Results are marked in red and CSV data is also displayed.

1
00:06:43.000 --> 00:06:46.000
The demonstration is now over.

1
00:06:46.000 --> 00:06:52.000
What you have seen is how independent Web resources can become rich, interactive map layers —

1
00:06:52.000 --> 00:06:57.000
and how those layers can be processed together in the browser.

1
00:06:57.000 --> 00:07:00.000
This is what Hyper Layering makes possible.