“I am at a foreign airport for the first time. I nervously look around for the check-in area, and then at the last minute I learn that there’s been a gate change. I foolishly rush off in what I think is the right direction, but I find myself back in the shopping area instead of the check-in area.”
“As I’m a foreigner here, I try to connect to the WLAN with my smartphone, but I mistype, so it doesn’t work. Another person looks at me sympathetically and shows me his HTC, which has an exact map of the airport. He asks if he can help me find something. My first thought is, he’s using Google Maps online. But he’s not. He says, ‘I don’t want anybody automatically capturing data without my knowledge about what I just bought in which store. So I use special OSM maps that I can conveniently download at home. OSM is an accurate map that is updated daily and that I can use offline. It has everything I need for a trip, including information about the airports I will be in’.”
The OpenStreetMap (OSM) project is not new to anyone in the geodata industry. First ignored, then ridiculed and criticised, but in a few years time this community mapping project has become a giant. As per the, ‘Global Digital Map Market 2014-2018’, OSM has become a serious (and probably long-term) rival to Google Maps.
Considering the success criteria of OSM, it can be said that the shared benefits offered by OSM — its openness, and the ease with which it lets even newbies capture data, as well as the fact that the license is free yet the software can be freely used commercially — are the key factors in motivating the OSM mappers, who now number over 2 million.
With the dynamic growth of the community, the original OSM approach of capturing streets, paths, rivers, borders, addresses, buildings, points of interest, and additional information such as opening hours, has been fully implemented in more and more areas.
But the OSM members plan to keep mapping and expanding the project. Future developments fall into two major trends. One is, of course, the mapping of areas is not yet completely captured. Mappers map more remote areas using aerial photographs. This is called ‘couch mapping’ in OSM language, and is limited to items that are clearly visible from the air like major roads, rivers, and prominent buildings. Couch mappers also visit other target areas and take part in projects such as Humanitarian Mapping.
Mappers also try to include new topics in the existing OSM definition by writing new proposals that expand the definition of a map. This includes simple things like including fire hydrants in maps (a point described as emergency=fire hydrant) and more complex topics like simple 3D modelling, micromapping, and indoor mapping.
The introduction of the ‘Simple 3D Building Definition (S3DB)’ pioneered this development. Just one year after the S3DB was set up, some 370,000 3D building models were made. Parallel to this, interesting projects came into being like Kendzi3D, OSM2World, Opensciencemap, and the commercial F4-maps.
Indoor map of an airport
In the community, ‘micromapping’ refers to the trend to include more and more detail and precision in traditional maps, right down to the precise mapping of road surfaces. The indoor mapping concept opens up even more interesting possibilities for the OSM project. The idea itself is simple. Now that the world outside a building has been precisely mapped, it’s time to add the interior structures of important public buildings to the map.
The first test tools for showing such a map have existed since early 2013. Since 3D mapping, which uses a different approach from indoor mapping, also emerged at the same time, an effort is now underway to unite the two trends through a new, complex definition. The new definition needs to provide the following postulates:
- Reverse compatibility with conventional OSM 2D map rendering (especially top-view renderings of buildings).
- Simplicity and obvious for OSM mappers.
- Indoor routing.
- S3DB compatibility.
- Selection of outdoor/indoor content for real-time rendering.
- Semi-automatic import of existing 3D models made by architects and civil engineers.
- True 3D rendering of whole structures.
- Depending on building level.
- Compatibility with Open Geospatial Consortium indoor definition.
New OSM proposal for indoor mapping
The first OSM indoor approach defines the building and indoor structure separately as shown in Figure 1. The building outline, attributed in OSM as building=yes, has no points of commonality with the inside structure, in which all rooms are shown as surfaces. The attributes assigned to the surfaces are names and /or functions. The ways lead to the points of entry to a given room.
This makes it difficult to make an accurate advanced 3D rendering of the outside and inside structures of the building. A generalised 2D representation is possible, however.
The solution, disputed with several involved community members, was proposed for full 3D rendering.
All rooms will continue to be mapped as enclosed spaces with a name (2BS02, or Starbucks Coffee, toilet, customs office, etc.) but the rooms adjoin the outer walls. Internally, the line dividing the rooms is always on the centre axis of the construction wall. Walls depict the centre axes of rooms. All walls are described with height and width, and optionally with material, colour, or texture.
The mapper then captures the centre points of the openings and describes the width and height of the opening and balustrade. The opening can be defined as a door, a window, just an opening (a hole in the wall with a girder), or a niche (does not go all the way through). Special additional attributes such as opening:direction=* for a door or the geometric shape (window=round) together with min_ height=*, material=*, colour=* and name=* (e.g. sign on door) give a precise description that can be used to generate an architectural drawing or full 3D rendering.
Once a floor is done, the user draws a ceiling which naturally has openings, for example for staircases and galleries. A trained construction engineer, architect, or CAD draftsperson will immediately recognise familiar elements from the typical 3D design tools used in the industry.
Without going into more detail, it will just be noted here that construction elements like ramps, staircases, columns, roof surfaces, roof windows, bays, dormers, and chimneys can naturally be described in this concept.
As with all other developments in OSM, such buildings will no doubt be added step by step. A voice from the OSM Forum: “The building needs to be there first. Just doing the interior (as accurately as you can) is already a step. Then an ‘architect’ comes along and dimensions it to the millimeter. The next guy has fun entering the 3D (simple buildings tags) and then somebody else does it even more precisely.”
Routing through the building is something that the architect knows nothing about. The most common approach is routing with a graph net, as is also used in the OSM community. So an OSM user would see the top priority as entering all possible routes in the building as a graph net for each floor.
The entrances to the building, which can be described as entrance=yes / main / emergency / exit /service, play the most important role in connecting the routing inside the building with the conventional routing outside it. Connectors between levels like stairways, ramps, escalators, and elevators get special designations to differentiate up from down
In large public buildings, in particular, indoor maps can help visitors find their way quickly and reliably. Examples include large hospitals, airports, universities, shopping malls, museums, etc.
If the new OSM modelling approach described here is adopted, the OSM map, as seen by the stranger at the airport, can look more realistic but without disturbing details and three-dimensional, like in simple 3D computer games. The OpenStreetMap approach offers numerous possibilities for the future of the geodata industry like:
- Universal navigation systems: Detailed descriptions of parking lots including headroom, escape routes, barriers, ticket machines, elevators, emergency telephones, opening hours, toilets etc.; train stations; airport navigation; and changing exhibitions in exhibition halls.
- Emergency and rescue maps in apps: In conjunction with location-based services, these kinds of maps can be used for temporary events. Temporary barriers, emergency exits, and emergency meeting points can be entered quickly, made visible on the app, and then removed after the event. The safety of escape routes out of buildings can be checked.
- Scientific analyses and simulations: Smog, smoke, and fire propagation through buildings or entire building complexes, earthquake safety etc.The creation of a universal, open-source information that is updated daily and where everyone can contribute knowledge will open up entirely new business opportunities for many companies. The French national railway SNCF is already working with the OSM community, for example, and the Yellow Pages in Poland use OSM in the conventional manner. So the future looks interesting for innovations in the geodata field. As a frequent flyer, I’m looking forward to it.
New Business Development, Elektrobit — Automotive