The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-4/W1, 2016

International Conference on Geomatic and Geospatial Technology (GGT) 2016, 3–5 October 2016, Kuala Lumpur, Malaysia

GEOPACKAGE DATA FORMAT FOR COLLABORATIVE MAPPING OF GEOSPATIAL

DATA IN LIMITED NETWORK ENVIRONMENTS
M. H. Rashidan, I. A. Musliman , A. A. Rahman

Department of Geoinformation, Faculty of Geoinformation and Real Estate, Universiti Teknologi Malaysia, 81310, Johor,
Malaysia - (ivinamri, alias)@utm.my

KEY WORDS: Collaborative Mapping, GeoPackage, Mobile GIS, Field Data Collector

ABSTRACT:

With the growth of technology in earth and space science informatics has led to the revolution in a wide range of geospatial practice.
Nowadays collaborative mapping has become a new hot spot, following mobile and web GIS. This paper explores the potential use
of GeoPackage for collaborative mapping of geospatial data in limited network environments. GeoPackage is a data format that
open-standard, platform-independent, portable, and self-describing. This paper focus on the implementation of GeoPackage in mobile
application for field data collection. A mobile application was developed that implements the GeoPackage data format as an internal
database to provide support for offline mapping. The developed mobile application demonstrates that vector and raster data can be
stored in a single data format, which reduces the device storage consumption. The details of how GeoPackage data contribute to mobile
GIS to achieve collaborative mapping in limited network environments are discussed. The findings show that the GeoPackage data
format has great potential to improve existing mobile GIS applications.

1. INTRODUCTION Many mobile GIS (proprietary or open source) were developed

by large group of companies and non-profit agencies offer mobile
1.1 Collaborative Mapping applications which support for outdoor mapping works. Field
data collector utilize such application to collect geospatial data
Collaborative mapping is an initiative to collectively create mod- for asset mapping, management, or monitoring. Some of the
els of real-world features online, then allow users to access and developed mobile applications comes with an advanced features
use to virtually annotate locations in space. It is known as a meth- such as ability to access multiple hosting database in real-time,
ods of generating content on the Internet, which involves contri- supports for online and offline editing, and integrates with ad-
butions from group of individuals or entities, and can take sev- vanced positioning technology. However, mobile GIS still faces
eral distinct forms. With the growth of technology for creating challenges in certain aspects like multiple datasets are stored as
and sharing maps, collaborative mapping have become a good individual file format in mobile device, thus lead to contain re-
initiation to incorporate new applications particularly in spatial dundant geospatial data which consuming large storage capacity.
information dissemination (Jokar Arsanjani and Vaz, 2015). The Also, the application requiring custom applications for data trans-
mapping applications vary depending on which feature the col- lation, replication and synchronization to enable different appli-
laborative edition takes place: whether on the map itself (shared cations to share the same view (OGC, 2015c).
surface), or on overlays to the map (feature overlaid). A very
simple collaborative mapping application would just plot users’ In mobile applications, the collaborative mapping takes place when
locations on maps (OpenLayers, Google Map and etc.). In this the map is generated collaboratively from multiple data sources
kind of application, the map itself is created collaboratively by (overlays on map). Locations and spatial information can be ac-
sharing a common surface, thus benefits to identify areas of share cessed by users through the map, and users able to query infor-
knowledge, reduce duplication of efforts, share knowledge of in- mation, insert new data, or update the map in real-time or offline.
terest, and develop new innovative ideas to address challenges. Before this, mobile GIS data mapping is highly depends on the
Thus, mapping of geospatial data nowadays have become easier Internet connection to operates, this is due to tight communica-
and social adapted. tion between server and client. Previous method used in offline
mapping for mobile application is through temporary datasets
At the moment, the emerging geospatial sharing technologies in created (retrieved from online database) and stored in local de-
earth and space science informatics has led to the transforming vice. Any changes that are applied to the datasets will be updates
practice in a wide array of disciplines, from the increasing use of to the main database when the Internet is accessible. This allows
desktop GIS, to distributed GIS (web), and ubiquitous GIS (mo- mapping can be carried out despite in the absence of an Internet
bile). At present, geospatial mobile applications are appearing connection.
at an increasingly rapid pace, such as in enhancements to uni-
fied critical communication suites (Business Wire, 2014), leak- This paper is organized as follows, in the next section will dis-
age management systems (MENA Report, 2013), real-time field cussed on constraints encountered in the mobile application to
surveys (Jeefoo, 2014), communication networks for disaster- achieve collaborative mapping. This is followed by introduc-
damaged areas (Youhei et al., 2014), smart eco-path finders for tion to OGC GeoPackage data format, its potential to contribu-
mobile users (Lwin and Murayama, 2013), time cost optimiza- tion in mobile application, and current status implementation.
tion for mobile GIS queries (Haifa et al., 2012) and so forth. Next, overview on collaborative mapping in utility mapping for
These patterns show mobile applications have become an impor- Malaysia scenario. Then the approach, framework and data in-
tant components in geospatial technologies. volved is described in details. Lastly, the findings are presented
and discussed.

This contribution has been peer-reviewed.

doi:10.5194/isprs-archives-XLII-4-W1-15-2016 15
 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-4/W1, 2016
International Conference on Geomatic and Geospatial Technology (GGT) 2016, 3–5 October 2016, Kuala Lumpur, Malaysia

1.2 Mapping in Mobile Environments In this paper, new data format called GeoPackage is utilized
and implemented in the mobile application development for
Mobile GIS, a cross-research field integrating modern mobile com- collabo- rative mapping of geospatial data (utility mapping).
puting techniques with geographic information science, has at- The details of how GeoPackage data contributes to mobile
tracted more and more researchers in exploring and developing GIS to archive collaborative mapping in limited network
new practices in spatial mobile application. It is built on mo- environments are dis- cussed and presented details. The
bile terminal environments of limited capacity, providing mov- advantages of GeoPackage overs the others data format are
ing, distributed and flexible mobile geographic information ser- also highlighted.
vices. Specifically, mobile GIS is based on embedded GIS inte-
gration with positioning technology, mobile communication tech- 2. OGC GEOPACKAGE DATA FORMAT
nology, and wireless Internet technology to expand the GIS ap-
plication, and improve GIS information collection, management Over 400 institutions are coordinated by the OGC in the
and analysis (Li, 2009). devel- opment of geospatial standards. During the last years,
two main trends have made disruptions in geospatial
Further development of technology in related fields has lead mo- applications: mobile and context sharing. People now have
bile GIS to be expanded its application areas in various disci- more mobile devices to support their work and personal lives.
plines. For instance many mobile applications were developed to Based on these trends new OGC data format called
support collaborative mapping, and some of these applications GeoPackage will enable greater geospa- tial data sharing on
are meant for data collection and public services. Previously, mobile devices (Singh et al., 2013). GeoPack- age is an open
mobile GIS for field data collection faced the problems of not standard, non-proprietary and platform-independent geospatial
working because of the absence of an Internet connection. When data format from the Open Geospatial Consortium (OGC,
access to the database server was interrupted, the mapping pro- 2015b). It is may become a modern alternative to formats
cess was delayed. However, many mobile GIS developers have like Spatial Data Transfer Standard (SDTS) and Shapefile.
developed an improved mobile application that offers the abil- GeoPack- age was introduced with the purpose of increasing
ity to operate in limited network environments. This is achieved cross-platform interoperability of geospatial applications and
through temporarily storing required datasets in the mobile de- web services since
vice. Any changes and updates made to the temporary datasets it is built on top of SQLite. Thus it enlarges the potential
will be sent back to the database server for updates. mar- ket for such applications and services, particularly in
resource- constrained computing environments. Furthermore,
Leaders in GIS industry are actively involved in research of mo- this format was designed and prototyped following a multi-
bile GIS and have launched related products on the data collec- year, open pro- cess of requirements testing and public input to
tion which support for offline mapping. Among them are ES- facilitate widespread adoption and use of a single file format
RIs ArcPad - a mobile field mapping and data collection software by both commercial and open-source software applications, on
designed for GIS professionals; WebMapSolutions - has experi- enterprise production plat- forms as well as on mobile hand-
enced building offline mobile GIS apps for over 4 years; GIS2GO held devices (OGC, 2015b).
- offline GIS app on mobile tablet; Mappt - an agile data capture
and management tool used on Android platform with advanced This data format is a single file format with a performance
map caching; ASTECs Gislet - an online and offline mobile GIS like a spatial database (implement a database schema) along
for data collection; MB&Gs MobileMap - a mobile field data col- with the convenience of a file-based dataset that can be
lection with offline GIS capabilities; and many more. emailed, shared on a flash drive or burned onto a digital disc.
In addition, this format efficiently stores vector geospatial
In collaborative mapping, problem arise when several geospatial features and/or tile ma- trix sets of earth imagery and raster
datasets are needed to assist the mapping activity, these datasets maps at various scales. It can be directly used in cross-
are downloaded and stored temporarily in the mobile device. This platform geospatial applications, which means it has the
will introduced problems like application to be slowed, hanged or ability to access and update data in a native storage format
crashed. The reasons are mobile applications are not design in- without intermediate format translations. Adding GeoPackage
tended to handle the massive spatial data processing request and support to an application can be done by an SQL-savvy
large amount of datasets. A good mobile application for collabo- programmer. Likewise, this data format guarantees the data
rative mapping should be lightweight and able to perform spatial model and dataset integrity, and is particularly useful on
process fluently. Therefore, even the current mobile applications mobile devices in communications environments with limited
are supported for offline editing, it is still limited to a certain in- connectivity and bandwidth.
stances.
2.1 Contributes to mobile application
The other issue in mobile applications are some geospatial data
are stored in individual file format. This will introduces to data Traditionally, mobile GIS applications required various types of
duplication and redundancy at the same time consuming large data (raster and vector) have to keep the data in an individual
storage capacity. For example, sometime mapping works requires format. This caused limitation in storage capacity and device
access to wide range of spatial data (vector and raster), which performance. Furthermore, only several mobile GIS applications
may consisting several types of spatial data formats. These data can be operated in both online and offline environment, and de-
formats have their own use and advantage, and some of them are velopers need to choose whether to opt for online (web-like) or
not suitable for mobile application. Currently the data formats offline (desktop-like) uses. Through GeoPackage, implementa-
that are likely to be used in mobile applications are Shapefile, tion for online and offline are possible and easier since the for-
Keyhole Markup Language (KML or KMZ), AutoCAD DXF, mat is a database container. For example, Adams and Suykens
Spatial Data Transfer Standard (SDTS), MDB Geodatabase (2013) studied the integration of geospatial mapping on mobile
(ES- RIs ArcGIS), GeoJSON, GeoCSV, MBTiles and embedded system which intended to increase situational aware-
JPEG2000. Some of these data formats are open standard, ness for emergency support. They stated their support for the
and some of them are proprietary data format. OGC GeoPackage standardization effort,

This contribution has been peer-reviewed.

doi:10.5194/isprs-archives-XLII-4-W1-15-2016 16
 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-4/W1, 2016
International Conference on Geomatic and Geospatial Technology (GGT) 2016, 3–5 October 2016, Kuala Lumpur, Malaysia

because this data for- mat will improve the existing system
since the system is used to deploy and store critical data, such as
aerial imagery and building plans, on the emergency workers
mobile devices. The advan- tages of using GeoPackage for
mobile applications is pointed as follows:
”For other field operations, involving dismounted soldiers or bor-
der patrol agents for example, this data format (GeoPackage) will
enable interoperable access to satellite, environment, and other
geospatial data in both online and offline environments”. (Adams
and Suykens, 2013)

Furthermore, GeoPackage can be implemented in web-like mo-

bile GIS to improve the functionality and capability of an applica-
tion. As Mas et al., (2013) has reported in their study of Geospa-
tial Web Services and GIS applications, by using GeoPackage, all
geospatial data can be encapsulated into a single file format, and
thus can be used in a disconnected environment. GeoPackage
can make web GIS applications possible works in offline envi-
ronments (client side) when access to the Internet is limited. This
data format has shown a huge potential to improve in the way cur-
rent mobile GIS application deployment and implementation.
Figure 1. GeoPackage tables overview.(Source: OGC GeoPack-
2.2 Current status and implementation age Encoding Standard)
On 13 February 2014, the Open Geospatial Consortium adopted At present, the subsurface utility data is acquired using Ground
the OGC GeoPackage (GPKG) Encoding Standard. The GeoPack- Penetrating Radar (GPR) and Pipe Cable Locator (PCL). Based
age standard will make it much easier to exchange and share on the current workflow shown in Figure 2, the utility mapping
geospatial (or location) information across different devices, ap- works were managed by two survey teams; first team using GPR
plications and web services. There is an initial charter of the and PCL to detect subsurface objects, and the second team using
Standards Working Group (SWG) for the GeoPackage, the pur- GPS to collect position (coordinates). Problem arise during the
pose of which is to produce a version 1.0.0 implementation stan- data updating and merging at desk office since the real situation
dard. This SWG is focused on defining and documenting a new of collected underground utility objects were presented slightly
OGC GeoPackage candidate standard based on the specific con- different from the field, and this introduced inaccuracies in the
tributions of existing work as a starting point (OGC, 2015b). The database (due to re-digitizing and point-to-point joining process
current version of GeoPackage encoding standard is 1.0.2, which by desk officer).
uses corrigenda process to make correction to the standard by
tracking issues and relying on public feedback (OGC, 2015a).
The first corrigendum was focused on the features portion of the
standard, and SWG is in the process of producing a second cor-
rigendum that will focus on tiles. It is believe this second corri-
gendum will further improve the readability of the standard and
reduce the risk of misinterpretation.

Moreover, this data format is a RDBMS container store with

SQL access for simple features with geometry to manage (cre-
ate, update, delete as well as search and retrieve) both geospa-
tial foundation data for multiple types of features, and newly col-
lected feature observation data. The standard defined that every
GeoPackage file will have three types of tables; required meta-
data, optional metadata and optional data as shown in Figure 1.

2.3 Collaborative mapping in utility mapping Malaysia sce-

nario
Figure 2. Utility mapping workflow.
In Malaysia, the Department of Survey and Mapping Malaysia
(DSMM) is the party that responsible in acquiring utility data na-
To overcome this problems, field-to-finish mobile GIS mapping
tionwide. The Cabinet of Parliament has decided that the DSMM
system was developed. The system was designed to provide func-
shall undertake the responsibility of maintaining a repository of
tionality for field surveyors to access online databases in real-
all underground utility data, apart from those kept by the vari-
time; update, query and edit features offline; and acquires posi-
ous utility agencies (DSMM, 2006). Maintaining reliable under-
tion through Real Time Kinematic (RTK) positioning. Further-
ground utility mapping information is an important task in the
more, the developed system implements OGC GeoPackage data
development, maintenance and upgrading of underground utility
format to support offline data editing, display and manage. All re-
infrastructure. Taking into account the increasing instances of
trieved data (vector and raster) from hosting databases are stored
catastrophic damages of underground utilities and disruption of
into a single file format (GeoPackage) in mobile device. This sin-
existing utility services resulting from excavation works, has led
gle file format is act as temporary database in mobile application.
to the establishment of centralized utility database.

This contribution has been peer-reviewed.

doi:10.5194/isprs-archives-XLII-4-W1-15-2016 17
 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-4/W1, 2016
International Conference on Geomatic and Geospatial Technology (GGT) 2016, 3–5 October 2016, Kuala Lumpur, Malaysia

is designed to be deployed in high spec tablet to provide seamless

Therefore, this paper explores the potential uses of GeoPackage and ubiquitous mobile application. Open source libraries are used
in collaborative mapping of geospatial data in limited network in the application development to avoid costs and give flexibility
environments using mobile GIS as the platform. in application design. Maps rendering and controls in the applica-
tion is backed by SharpMap library, and new library is written to
add additional support for reading and writing GeoPackage data
3. METHODOLOGY format. The process of data translation from database server to
GeoPackage and vice versa are handle by a pipeline process.
3.1 System application framework
This process is one of the main features in the mobile appli-
The concept used in this system application is based on the utility cation. The process will translate the spatial data from Post-
mapping process (outdoor). The system is designed and devel- greSQL (PostGIS) and MySQL (spatial) into GeoPackage for-
oped combining two major components, namely database server mat (for vector), and from FTP server (GeoTiff) into GeoPack-
and mobile application. Database server is a component which age format (for raster). The pipeline process is designed and
acts as a medium to provide access to the spatial data used by the developed integrated with Geospatial Data Abstraction Library
mobile application. While the mobile application will access the (GDAL). GDAL is a translator library for raster and vector geospa-
data from the database server through the Internet access. This tial data formats that is released under an X/MIT style Open
process involves communication between the database server and Source license by the Open Source Geospatial Foundation. As
the mobile application. In particular, the mapping work involved a library, it presents a single raster abstract data model and vec-
in this process can be categorized into two parts, online and of- tor abstract data model to the calling application for all supported
fline. The purpose of providing offline option is because occa- formats. It also comes with a variety of useful command-line util-
sionally mapping process involve in area with limited network ities for data translation and processing (GDAL, 2015).
environments. The online part as shown in Figure 3 displays
communications between the database server and the mobile ap- Besides that, the developed mobile application was integrated
plication. The mobile application will send spatial requests to the with external GPS receiver to stream GPS data. To achieve higher
server, and the server will process the requests and send the re- levels of accuracy and precision, positioning augmentation cor-
quired datasets to the mobile application. rection via wireless communication integration between GPS
and IskandarNet (a scientific research Continuously Operating
For offline part as seen in Figure 3, the obtained required datasets Reference Station (CORS) network) via NTRIP system was im-
(from database server) are translated into GeoPackage data for- plemented. This is because standard GPS receiver have limited
mat. This is because GeoPackage is act as intermediate data for- use and application due to the impact of random and systematic
mat to be used in offline mapping. The mobile application will errors associated with real-time autonomous positioning such as
reads, access and displays the data in the form of map. Thus user atmospheric effects, broadcast ephemeris errors, multipath and
access and view existing spatial data in mobile application. This receiver noise.
is intended to facilitate the field data collector (user) to identify
the surrounding mapping area quickly. A device called GPR or 3.3 Data
PCL is used to scan and track the position of subsurface utility
object. After detected, field data collector will mark the position This section describes the number, type and source of data in-
of the utility objects, and coordinate of the detected objects will put used in this study for mobile utility mapping works. The
be taken through RTK-GPS positioning. The readings obtained data input comes from three main sources, namely PostgreSQL,
from the process (coordinate observations) will be recorded and MySQL and FTP server as can be seen in Table 1. The sources
updated into the mobile application. At this time, the process of (database server) provide five spatial layers to be used in the util-
updating and editing occurs on the temporary datasets (GeoPack- ity mapping process. They are basemap, river, pipeline, road and
age) in mobile device. This process will be continues until all the satellite. Basemap layer provides information about land parcel,
identified utility objects are scanned. The final process is all new while pipeline layer is used for utility mapping purpose. Satellite
collected utility data will be sent to the database server for up- layer is a raster data which provides satellite imagery around the
dating. The mobile application will send requests to the database mapping area. While river and road layers provides information
servers for data append (the new data will be uploaded). respectively as the name implies. Another source of data input is
comes from GPS data (coordinates) which is used for mapping
utility objects.

Table 1. List of layers from multiple data sources.

Figure 3. Utility mapping concept used in this study.
4. RESULT
3.2 Mobile application development
The mapping process begins with the required datasets are down-
A mobile GIS application was developed with the purpose to as- loaded from the online database servers through the mobile ap-
sist the utility mapping works. Its provides access to the hosting plication. The required datasets are basemap, river, pipeline, road
spatial database (PostgreSQL and MySQL) and FTP server. The and satellite layers. These downloaded datasets are going through
program is written in C# and based on the Net 4.5 framework. It

This contribution has been peer-reviewed.

doi:10.5194/isprs-archives-XLII-4-W1-15-2016 18
 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-4/W1, 2016
International Conference on Geomatic and Geospatial Technology (GGT) 2016, 3–5 October 2016, Kuala Lumpur, Malaysia

the process of data translations (pipeline). During the process, all have additional tables that belong to the layer. For example, a
data are translated into GeoPackage data format and encapsulated layer pipeline have five tables belong to it, they are pipeline,
into a single file named database.gpkg. The database.gpkg is act rtree pipeline geom, rtree pipeline geom node, and others as shown
as a temporary database in mobile application during the map- in Figure 6. The tables starting with the name rtree is a spatial in-
ping works. Its contained several layers inside, which consisting dex table that created to accelerate the spatial queries. While the
of vector and raster data. Figure 4 illustrates the mobile applica- raster table (satellite) is only have one table, named satellite. The
tions map view showing multiple layers being mapped collabo- others tables (gpkg spatial ref sys, gpkg contents, gpkg extensions,
ratively offline. The map shows data from layers basemap, river, gpkg geometry columns, gpkg metadata, gpkg metadata reference,
road and pipeline. These layers are accessed from database.gpkg gpkg tile matrix and gpkg tile matrix set) are the default tables
located in the device storage, and access to online databases does in the GeoPackage standard. Each of these tables have their own
not become a necessity. Besides, user also able to identify, pan, uses and purposes.
and zoom (in and out) the maps contents using the available map
toolbars. Meanwhile, Figure 5 shows the raster layer (satellite)
being mapped collaboratively with others layers in the mobile
application.

Figure 4. The mobile mapping application main interface shows

multiple layers being mapped collaboratively offline.

The developed mobile application demonstrate the ability to mapped

geospatial data (vector and raster) in the absence of Internet con-
nection. This is due to implementation of GeoPackage as internal
Figure 6. The data organization structure in GeoPackage for data
database in the mobile application. The data format extends the
capability of the developed mobile GIS application. Hence using integration of vector and raster (database.gpkg).
GeoPackage in mobile GIS has caused the mapping process to
become easier and efficient. For vector data, the geometry data with or without optional eleva-
tion (Z) and/or measure (M) is stored in SQL BLOBs. While for
raster data, each tiles data are stored in BLOB data type. Besides,
list of the available features in GeoPackage and its data type in-
formation can be obtained by accessing the gpkg contents table.
GeoPackage can be directly used to access and update data in a
native storage format through the use of SQL command.

4.2 Simple and lightweight

The GeoPackage file (database.gpkg) which used to store trans-

lated datasets from database server has shown its lightweight in
storing multiple datasets. Figure 7 shows the comparison in term
of file size between database.gpkg and satellite.tif. Its shows
that database.gpkg have a file size of 3.05 megabyte, while satel-
Figure 5. The mobile mapping application main interface shows lite.tif have a file size of 19.8 megabyte. What interesting here is
database.gpkg contains data consisting of basemap, river, pipeline,
raster layer (satellite) being mapped collaboratively with others
road and also includes data from the satellite.tif. Plus the file
layers using data from database.gpkg.
size is smaller than the satellite.tif in significant difference. This
shows GeoPackage is good in storing large size of imagery data.
4.1 Data organization structure
It should be noted that more comprehensive study is needed to
study the capability of GeoPackage in storing others raster data
The data organization structure of GeoPackage as a single file
format.
format can be seen in Figure 6 (vector and raster). The basemap,
river, and pipeline data come from the PostgreSQL database, while
4.3 Utility mapping
road data comes from the MySQL database. Raster data (satel-
lite) comes from FTP server. These layers are stored as data ta-
Utility mapping process begins with the identification of utility
bles in GeoPackage. A vector layer stored in GeoPackage may
object to be mapped, and then GPR / PCL is used to detect the

This contribution has been peer-reviewed.

doi:10.5194/isprs-archives-XLII-4-W1-15-2016 19
 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-4/W1, 2016
International Conference on Geomatic and Geospatial Technology (GGT) 2016, 3–5 October 2016, Kuala Lumpur, Malaysia

tinue to find new and innovative uses for it. Some of them will be
mobile-based and some will not.

It is know that Shapefile has become prominent data format in

the last few years. Currently shapefile is challenged due to sev-
Figure 7. Comparison of two data format in file size. Left: eral reasons. The reasons are maximum length of dbf field names
GeoPackage data format containing vectors and raster data. is restricted to 10 characters; filenames should be no longer than
Right: GeoTiff data format (satellite image). 8 chars; no layer name (has to be derived from the filename);
maximum file size is restricted to 2GB (dbf); four and more clut-
position and depth of the utility object. After that the identi- tered files (.shp,.dbf.,.shx,.prj, encoding, other indexes, etc.); only
fied utility object is marked for the purpose of coordinates ob- MultiPoint and no MultiLinestring, MultiPolygon, GeometryCol-
servations using GPS receiver. While in the mobile application, lection and Arcs; only one geometry per table; no topology; and
the coordinates of the utility object with high accuracy readings no data schema (reference). The current possible data formats to
is used and added to the mobile application via data editor in- replace Shapefile would be the GeoPackage or GeoJSON.
terface. The added coordinates will be shown on the map as
a cross-sign (cyan color) as seen in Figure 8. When the map- GeoJSON is an interchange geospatial data format based on JSON
ping is done, field data collector can instruct the mobile appli- (Javascript Object Notation). It supports the following geometry
cation to save the update. The application will shows the latest types: Point, LineString, Polygon, MultiPoint, MultiLineString,
data update via the map viewer. The mobile application write and MultiPolygon, and very popular in web mapping applica-
the geometry data into GeoPackage file using the standard SQL tions. Despite GeoPackage and GeoJSON formats overcome
ST_GeomFromText() function. New data are updated to the many of Shapefiles limitations. GeoPackage is more complex
tem- porary database (database.gpkg). This updated datasets are than Geo- JSON but it also provides a number of additional
later used to update the main database when the device is capabilities that some scenarios demand. As the capabilities
connected to an Internet. of a database, ac- cessing and manipulating data is better than
GeoJSON. For in- stances, spatial analysis can be done at data
level, through simple or complex SQL spatial queries with
extensions (e.g. ST_Area, ST_Buffer, ST_Centroid,
ST_Contains, ST_Disjoint and etc.).

In this study the raster data (satellite imagery) is used to pro-

vide images of reality. Normally MBTiles format is used to store
tiles data and have been seen widely used in web map application
to efficiently stores millions of tiles in a single SQLite database.
The format is portable, single-file solution for storing and serv-
ing web maps. GeoPackage differs from MBTiles in a few ways.
The MBTiles only supports tiles in the Spherical Mercator projec-
tion, while this is the default in GeoPackage, other projections are
supported through an extension. MBTiles only supports a single
tileset in a file, while GeoPackage supports multiple tile sets in
a single file, as well as feature data and metadata (OGC, 2015a).
Figure 8. Newly added coordinates (points) are shown as X sign This indicates that GeoPackage is also efficient stores tile matrix
(cyan color). in a single file based.

The concept used in this study has put GeoPackage as interme-

5. DISCUSSION AND CONCLUSION diate data format to supports collaborative mapping in limited
network environments. Because of this, the data translation pro-
The developed mobile application for outdoor mapping have shown cess is a must to enable application to operates in such condi-
great advantages especially in spatial data management and col- tions. Thus, further study is needed to looks on how translation
laborative mapping of geospatial features in mobile platform. Spa- process affected the original data model and data integrity. Issue
tial data can now be managed, accessed, edited and updated in to what extent the data model is preserved and maintained from
mobile device seamlessly without worries about the compatibil- the process of data translation from database (e.g. PostgreSQL)
ity, suitability and reliability of data format. GeoPackage has to GeoPackage, and vice versa. Whether the translation affected
comes to fulfil the needs in mobile application, and this format changes in data types (from integer to bigint, float to double, var-
seem to encourage the process of data sharing which lead to greater char to string and etc.) and data precision (0.01, 00.01 and etc.)
mapping activities. GeoPackage is simply a schema for data needs to be clarified. If this happened this could be loss to the
stored in an SQLite database and it is not just a standard for mo- data during the format translation.
bile.
Many GIS software and programs have support reading and/or
While mobile is clearly a major thrust of GeoPackage, it is not writing GeoPackage. Among them are QGIS, ArcGIS 10.3 and
the only one. The great thing about SQLite is that it can be ArcGIS Pro, GDAL, GeoServer, GeoTools, OpenJUMP PLUS,
used in nearly every modern computing environment from a Skyline, Luciad, Envitia MapLink, SpatiaLite, FME Desktop and
mobile device to the desktop up to the cloud. Because GeoPack- FME Server, Terrago and others. There is also versions for mo-
age is so flexible, users are envision it being used in a wide va- bile Android (Luciad GeoPackage Viewer).
riety of ways. GeoPackage have been foreseen being a partial
replacement for the Shapefile format. Several organizations are In conclusion, the use of GeoPackage for collaborative mapping
exploring ways to distribute geospatial data assets in the format. of geospatial data extend the ability to support on-site spatial data
As organizations continue to adopt GeoPackage, they will con- access, editing, capturing and updating, mainly for outdoor or in-

This contribution has been peer-reviewed.

doi:10.5194/isprs-archives-XLII-4-W1-15-2016 20
 The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XLII-4/W1, 2016
International Conference on Geomatic and Geospatial Technology (GGT) 2016, 3–5 October 2016, Kuala Lumpur, Malaysia

door field works processes whereas environment network is lim- Mas, J., Diaz, P., Riverola, A., Daz, D., and Pons, X., 2013. Ex-
ited. This data format has a standard APIs for access and man- changing the Status between Clients of Geospatial Web Services
agement which increase interoperability and result consistency. and GIS applications using Atom. In Proceedings of the Inter-
In addition the come with the support for vector and raster data. national MultiConference of Engineers and Computer Scientists
It has shown great potential as a robust spatial database in mo- (Vol. 1).
bile GIS application (outdoor mapping). This format will en-
ables greater geospatial data sharing due to its ability to fulfil MENA Report, 2013. Corporate GIS, mobile GIS and leakage
current needs in ubiquitous GIS development. Its interoperabil- management systems. Retrieved from https://vpn.utm.my/docview/
ity in cross-platform geospatial applications, efficiency in storing 1432351908.
multiple vectors and tile matrix in a single file, ability to directly
access and update without intermediate format translations, its OGC, 2015a. GeoPackage, an Open Format for Geospatial Infor-
lightweight and minimal implementation, its open and standards- mation. Retrieved from http://www.geopackage.org/
based non-proprietary data format, its capability for online and
offline modes, and its database-like capabilities will enlarge po- OGC, 2015b. GeoPackage Standards Working Group (SWG).
tential market values for mobile GIS deployments. Retrieved from http://www.opengeospatial.org/projects/groups/geo
packageswg.

ACKNOWLEDGEMENTS OGC, 2015c. GeoPackage: Information and Data Models, Open

Geospatial Consortium network. Retrieved from http://www.ogc
The authors would like to express their gratitude to the research network.net/geopackage.
grant sponsor; the Ministry of Science, Technology and Innova-
tion (MOSTI), Malaysia via ScienceFund grant (SF 04-01-06- Singh, R., and Bermudez, L., E., 2013. Emerging Geospatial
SF1187) and Universiti Teknologi Malaysia (Grant No. 01G32) Sharing Technologies in Earth and Space Science Informatics. In
for their support in conducting the research. Gratitude to research AGU Fall Meeting Abstracts, vol. 1, pp. 08.
grant sponsors.
Youhei, K., Ashraf, M., D., Bert, V., Masahiro, H., Takeshi, S.,
Itaru, Ki., Hajime, N., and Kento, I., 2014. Using GIS to develop
REFERENCES a mobile communications network for disaster-damaged areas,
International Journal of Digital Earth, 7:4, pp. 279-293.
Adams, B., and Suykens, F., 2013. Astute: Increased Situational
Awareness through proactive decision support and adaptive map-
centric user interfaces. In Intelligence and Security Informatics
Conference (EISIC), European IEEE. pp. 289-293.

Business Wire, 2014. Everbridge adds mobile, GIS enhance-

ments to unified critical communication suite. Retrieved from
https://vpn.utm.my/docview/1552692341.

DSMM, 2006. Standard Guideline for Underground Utility Map-

ping. National Mapping and Spatial Data Committee Malaysia,
the Department of Survey and Mapping Malaysia.

GDAL, 2015. Geospatial Data Abstraction Library http://www.

gdal.org/.

Haifa, E., E., and Souheil, K., 2012. A time cost optimization
for similar scenarios mobile GIS queries, Journal of Visual Lan-
guages and Computing, vol. 23, Issue 5, pp. 249-266.

Jeefoo, P., 2014. Real-time field survey using android-based in-

terface of mobile GIS, 5th International Conference on Informa-
tion Science and Applications.

Jokar Arsanjani, J., Vaz, E., 2015. An assessment of a collabora-

tive mapping approach for exploring land use patterns for several
European metropolises, International Journal of Applied Earth
Observation and Geoinformation, Volume 35, Part B, March 2015,
Pages 329-337, ISSN 0303-2434, http://dx.doi.org/10.1016/j.jag.
2014.09.009.

Li, B., 2009. Mobile Geographic Information System Technol-

ogy, Xian: Xi’an University of Electronic Science and Technol-
ogy Press, pp. 90-120.

Lwin, K., K., and Murayama, Y., 2013. Smart eco-path finder for
mobile GIS users. URISA Journal, 25(2), 5+.

This contribution has been peer-reviewed.

doi:10.5194/isprs-archives-XLII-4-W1-15-2016 21