GISdevelopment.net ---> GITA 2000 ---> Exploiting Field and Mobile Technologies

Mobile Geospatial Applications: A New GIS Paradigm

Tadeo H. Schultz
Geographic Information Technology, Inc. (GeoIT)
101 Inverness Drive East, Suite 130
Englewood, CO 80112, Usa
Phone (303) 708-9355 ext 119
Email : tschultz@geoit.com


Abstract
The availability of such technologies as GPS, wireless communication, PDAs and pen based systems coupled with compact GIS (Geographic Information Systems) and database components has created a fertile environment for a new breed of GIS applications. GIS can be the enabling paradigm that unifies this wide array of technology in a form palatable to a surprisingly varied and diverse consumer public.

Mobile Geospatial Applications will be the next wave in GIS as they fuse the power of GIS systems previously reserved for the desktop with the flexibility of compact PDAs and the convenience of GPS and wireless communications. This paper reviews some of the more popular forms that these mobile applications may take as well as the key technologies behind them. Applications models and examples are presented and Software frameworks, architectures and development strategies are discussed.

Introduction
Mobile Geospatial Applications can be thought of as ‘GIS on the go.’ They are GIS applications intended to be used in the field for the purpose of locating and cataloging resources or for assisting the mobile worker in a particular task that requires spatial information (vehicle navigation/tracking). The domain of mobile geospatial applications is an enabling technology that is synergistically combining the following leading edge technologies: Portable computing, GIS, GPS/GPS, wireless communications, lightweight databases, and distributed software components.

The ever-increasing push for interconnectivity and business intelligence in the field, coupled with the inherent capability of GIS systems as tools for information synthesis promises to make mobile GIS the ‘killer app’ of the next decade. The widespread availability and decreasing cost of such technologies as GPS and the wide array of choices in both communications and lightweight computing platforms have allowed mobile GIS to evolve from the limited vertical market to the horizontal mass market.

Bachground
In order to understand the current role of mobile GIS applications one must first look at three different trends:

The relentless push towards connectivity, interoperability and distributed information, the coming of age of the mobile worker, and the ever -increasing demand for geospatial information.

Connectivity and Interoperability
The web has become the ubiquitous neural backbone linking us all together. Such business practices as ‘just in time inventory’ has placed great demand for accurate and timely information. The relentless expansion of e-commerce is pushing new standards for security and the ability to effectively conduct business over the web. These capabilities in turn have created an even greater, more dynamic and freer marketplace. Simply having the connections is not enough. The IT landscape has become increasingly dependent on multi-tiered applications, distributed objects, distributed data, and intelligent agents thus making the dividing lines between data, software and network very diffuse.

Mobility
A recent white paper by IDC has identified the mobile worker, the proliferation of distributed web based apps, and the usage of intelligent wireless mobile computing devices as ‘the third paradigm of computer usage’ (IDC 1998). The explosive growth of mobile computing has truly been phenomenal. Some estimates for the number of mobile workers worldwide are as high as 108 million by the year 2001. Sales figures for handheld devices in the U.S. are expected to exceed 14.1 million in 1999. This figure does not include laptop or notebook computers. The phenomenal growth of the Internet and the World Wide Web is undoubtedly fueling this trend. Equally as provocative has been the rapid rise in wireless communications, and the accompanying services, i.e., digital and cellular voice and data, paging, email, fax, and Internet access.

Geospatial Information
GIS has steadily grown from its modest beginning in the 1960s as a mainframe based tool for land use planning to its current stature not only as the direct result of the ever decreasing $/MIPS ratio, but also because maps are a powerful tool for conveying useful information. Maps have existed since the Stone Age, predating the written word. They represent the quickest way of conveying spatial information and relationships non-verbally. GIS gives a virtual map with the additional capabilities of spatial queries and data edits at the click of a button. The mobile worker is dependent on accurate spatial information such as their current location, their destination, and the location of the assets with which they will be working.

Technology
The following is a brief review of the various technologies that makes mobile GIS applications possible. There are four major areas of emphasis, which include computer hardware, platform architecture, GPS, wireless communications, and software.

Computing Hardware
The ideal computing platform for mobile GIS is one that combines power, portability, and durability. Portable computing platforms in ever-decreasing size and cost and with ever-increasing performance profiles have made cost an irrelevant issue. The typical platform for mobile GIS tends to be a laptop computer, although smaller pen based systems and PDAs (personal digital assistants) are becoming more popular and are more ergonomically suited for mobile work.

PDAs and handheld devices – For mobile platforms the emphasis is now on size (referred to as the form factor), overall external design, ease of use and communications capabilities. PDAs, pen based systems, smart phones, and other hand held computing devices have a great potential in that they offer new user interface modalities that allow the form factor to continue to shrink. Laptops and portables are limited by the ergonomics of the keyboard. Also keyboard -based devices are difficult to use under certain conditions such as driving. A voice-activated interface is another desirable feature. PDAs and handheld devices are diskless and, in general, tend to be more rugged than laptop computers. GIS applications are inherently graphical and therefore well suited for pen-based systems.

Platform Architecture
Laptop based systems essentially duplicate the desktop environment (Windows NT/98, MacOs, and Linux). In terms of innovation and potential what is promising are the OS and architectures for the new hand held devices.

Palm OS and the Palm Architecture – 3COM has done an excellent job in designing the Palm architecture from the ground up and as a result has garnered extensive accolades and strong demand for the Palm platform. There are now over 10000 applications and 50000 registered software developers writing applications for the Palm Pilot. In addition, 3Com has agreements with Sony to incorporate the Palm OS into consumer electronic devices, and an agreement with Quallcom to incorporate Palm OS into its CDMA digital phone.

Microsoft Windows CE - Microsoft entered the market late in the PDA arena but its Windows CE OS for hand held computing is delivering the promise of a robust OS for handheld devices. It can run tailored down versions of Microsoft applications

Java – Java gave us interoperability across platforms thanks to the Virtual Machine technology. JINI, a networking Java based technology will allow all kinds of software and hardware components to coexist seamlessly on a network. Various players (Sun, IBM, Ericcson) are currently working on a variety of projects, including Java OS for phones, PDAs, and other devices.

Wireless communications
The ability to connect to a corporate Intranet or the World Wide Web from anywhere without the requirement of physical wires is what makes wireless communications so valuable and mobile GIS applications possible. Essentially wireless communications involve the usage of radio waves as the transport layer. Many possible choices exist ranging from dedicated wireless LANs to two-way wide area packet data networks. Table 1 summarizes wireless communications options.

Type Description Network/Technology
Dedicated wireless LANs Allows roaming in limited areas typically a building or campus while maintaining a wireless connection to the regular network Wired network is usually an Ethernet backbone.
Wireless circuit switched Data transmitted over existing cellular network, (session based)  
Two-way wide area packet data networks Available in all major metropolitan areas in US, Europe, and Australia ARDISS (DataTac Motorola) RAM (Mobitex)
CDPD – Cellular Digital Packet Data Overlays the regional cellular network, shares the network with voice. All the major cellular providers offer CDPD
GSM – Global System for Mobile Communication European digital cellular system primary purpose was to allow international roaming Requires a SIM (Subscriber Identity Module)
Paging (one-way /two-way) Available everywhere. Most common form of wireless. EMBARC
SMR – Specialized mobile radio Uses high powered mobile transmitter and one base transmitter for an entire metropolitan area. Vertical markets specialized application areas. Nextel (ESMR Motorola)
Satellite based digital networks Ideal for work in remote areas (oil and mineral exploration, research stations). High cost. IRIDIUM
HDR –High Data Rate To be offered soon. Data rate is comparable to cable modem 2megabytes/sec. Drawback is must have dedicated bandwidth. HDR (Qualcomm)

GPS/DGPS
The advent of GPS/DGPS has done more for Navigation and GIS than any other technology. GPS (Global Positioning System) was developed by the Department of Defense as a tool for accurate navigation and positioning. It utilizes 24 satellites whose orbits are accurately known and can act as reference points. GPS utilizes 3 dimensional triangulation by measuring the travel time between the satellites and the ground receivers (the GPS receivers) whose position on the surface of the earth is not known. Essentially if the receiver locks on to one satellite one can describe a sphere whose radius is the distance based on the travel time of the signal emitted by the satellite. The position of the receiver is somewhere on the surface of that sphere. If the receiver locks on to two satellites then an elliptical circle where the spheres intersect describes the potential solution. To narrow the solution to a point, 4 satellites are required, although in practice, 3 satellites could suffice since a two point solution would most likely result in one of the points lying beyond or below the surface of the earth. The solution is given in terms of latitude, longitude and altitude.

DGPS: Differential GPS is a technique for increasing the accuracy of GPS. Without DGPS the best measurements are at best
+/- 150 feet; with DGPS accuracy increases to within a couple of feet. DGPS works by subtracting out the effects of atmospheric and other transient effects by comparing the computed position of a well know reference point and comparing it to the actual position of the reference point. From this discrepancy a correction to the travel time can be computed which is then applied to our GPS measurements. The differential correction could then be applied in near real time if a communication link is present between the reference station and the GPS receiver, or as part of a post processing operation, if
The measurements are to be stored and used later.

Software Technology
The following reviews some of the software technologies that are making mobile GIS applications possible.

GIS Components – Typically GIS software includes functionality to display, query, manipulate and analyze spatial data and to link enterprise data to its spatial or geographic attributes. The advances in this area have been in the introduction of object/component based software which facilitates the development process. Mobile GIS applications have benefited in that developers can now pick and choose the functionality they require and need not bundle superfluous code in their mobile apps.

The following table lists the primary software vendors with component packages suitable for mobile application development.

Vendor Package Platform Description
ESRI MapObjects ActiveX GIS object collection includes geocoding, query and display, network analysis, GPS interface, web enabling, raster and image integration
MapInfo MapMarker, MapBasic ActiveX GIS tools for spatial query and display, geocoding and reverse geocoding
Autodesk MapGuide Java Web enabling components for a wide range of GIS data formats. Query and display
BlueMarble Geographic   ActiveX Spatial Query and Display


Database Components - Modern database technology is an integral part of any GIS. Traditionally GIS packages/systems included their own RDBMS, however, in recent years the trend has been towards interfacing and interoperability with major DBMS systems. In recent years mobile applications have been boosted by the appearance of small footprint RDBMS systems. These systems are configurable on a per application basis resulting in a DB memory fingerprint that is very small and capable of residing comfortably within an embedded or mobile system. In addition these systems include a synchronization and replication capability that allows mobile systems to be only sporadically connected thus realizing great savings in communications cost. The benefits of this technology within mobile applications are immense. Essentially this allows the enterprise database resources to be accessible within the realm of the mobile worker. The following reviews the offerings of the two major vendors:
  • Sybase - Invented the adaptable, small fingerprint database technology and is the leader in the field of databases enabled for mobile or embedded applications. Their product, Sybase SQL Anywhere Studio and Adaptive Server Anywhere, provide all of the aforementioned features. The memory fingerprint can be as small as 50K. They provide multi DBMS support. Application Platform can be Windows CE/NT/98, PalmOs, Psion Epoc. Language support includes Java, ActiveX. Enough tools are provided to make middleware unnecessary. Support is also provided for various wireless communications options.
  • Oracle – The database flagship has release a product called Oracle8i/Lite which, as the name implies, is a product falling within the Oracle 8 family. It has been extended to include configurable footprint, synchronization capabilities and a wide range of communication options.
OOD Technology - Component based software (ActiveX, Java/JavaBeans), distributed objects and the emerging standards (DCOM, CORBA) are revolutionizing software development. As more organizations adopt this technology many of today’s software development issues, i.e., interoperability problems, development cost, will become greatly diminished.

Application Models
A wide variety of mobile GIS applications are possible. Qualitatively the applications can be classified based on three indices: Measurement, Information, and Communication (Figure 1).


Measurement oriented applications are primarily concerned with measurements on the earth surface. They rely heavily on GPS and other measuring devices. Information oriented GIS applications concern themselves with the spatial relationships and the informational content of the data, such applications are more heavily dependant on database technology and enterprise datasets. An example would be AM/FM GIS. Communications-oriented applications emphasize the communications aspects. An example is a GIS-based Emergency Dispatch System. Another informal classification is based on areas of emphasis or application domains. The four major application areas are:

Geodesy and Land Survey GIS
These are measurement-oriented applications primarily concerned with the determination of distances, angles and coordinates on the earth surface. These systems are heavily dependant on DGPS and other telemetry devices, e.g., theodolites, total stations. “The impact of the Global Positioning System (GPS) on geodetic control surveys has been immense. In the past we have relied upon line-of-sight instrumentation to develop our coordinates. With GPS, ground station intervisibility is no longer required, and we can survey with much longer lines. Different instruments and survey techniques were used to measure horizontal and vertical coordinates, leading to two different networks with little overlap. GPS, on the other hand, is a three-dimensional system. “ (N.O.A.A. 1997).

These applications will use GIS for data storage and display. These systems place a premium on mobility. Database queries are less important than accurate computations. In terms of measurement precision these systems achieve centimeter level resolution. A vendor who is at the forefront in this application domain is Leica. They market software and instrumentation solutions that work seamlessly (Leica SKI, Leica LISCAD Plus).


Figure 2. A simplified DGPS /GIS land survey system

Mobile Land Use and Resource Management GIS
These applications emphasize field tabulation and classification of data. Typically, the users are in the environmental, or natural resource management (e.g. forestry) industry. These systems tend to emphasize statistical analysis of spatial data for the purpose of classification of land areas. Many of these systems also incorporate raster GIS. Here again DBMS play a lesser role. Traditionally, many of these systems are custom solutions developed either via university research or by government agencies (EPA,USGS, and NOAA). However, there also are private firms deploying these systems since their capabilities can be applied to a wide variety of problems

Vehicle Tracking, Navigation and Fleet Management GIS
These systems rely on network GIS datasets, e.g., road networks, and on GPS to perform their function: tracking the position of one or more vehicles with respect to maps and charts displayed on a GIS in near real time. They rely on GPS, communications, and also use of the GIS and DBMS. An example of this is in the case of routing, where given a destination address and a current location, the system will calculate the most efficient route through the road network for the vehicle to follow. This relies on GIS queries against the road network database and thus makes higher usage of the DBMS. Another interesting problem that arises in vehicle tracking is that the vehicle positional information returned by the GPS is sometimes more accurate than the underlying road maps used by the GIS. In these cases the position is snapped to the closest, most likely location on the network. Another interesting technique is know as ‘dead reckoning’ where additional sensors on the vehicle will allow a continuous readout of the vehicles position even in places where the GPS is unusable ‘urban canyons’ or in between GPS measurements.

AM/FM GIS
Utilities and telecommunications companies are beginning to make extensive use of mobile GIS technology. Because of corporate assets that are geographically distributed, AM/FM GIS applications tend to make extensive use of the database and use GIS datasets that exploit networks as well as thematic layers. Some of the applications are facilities information systems, outage management, and workforce management. Typically, AM/FM GIS is but a small part of a multitiered distributed enterprise information system (Figure 3). Some of the more interesting mobile GIS applications are coming out of this domain. For example:
  • Autodesk\ Mauii - A multitiered applications -based on MapGuide that runs the GIS display client and Oracle8iLite on the PalmPilot and can re-sync data with the corporate databases via wireless. The map displays allow for various levels of detail and attributes.
  • Cartesia – A wireless pen based mobile GIS product developed by hardware vendor Tadpole (know for UNIX laptops) and software vendor ILOG. The product uses a Java based GUI (ILOG Jview)
  • TUMSY – A proprietary system developed by Tokyo Gas Co., for workforce management. It utilizes GPS, GIS (query and display and data editing), and a wireless communications interface.

Figure 3. Simplified utility workforce management system with mobile GIS field application

Development Strategies
A simple strategy for developing efficient and usable mobile GIS applications consist of four points which are:
  • Simplified GUI – Spend time upfront with the users and design a simple GUI that is easy to use during field operations and conditions (e.g. driving, walking). Remember that the user may not have the time to focus completely and may have to use the application in conjunction with other tasks. Such an assumption that he or she will be free to type may be an erroneous assumption. Make sure buttons, icons and text, are easily visible and recognizable. Also, make sure that functions are not deeply nested. Users should not have to go down more than one level to get to all functions. Toolbars tend to be better choices than menu bars for these applications.
  • Restrict Functionality – Identify the scope of the functionality and restrict it to only those tasks that need be performed in the field. Typically, desktop GIS is feature and function rich since they are geared towards a wide audience and a wide range of problems. In the field, the mobile worker does not have the time and patience to run ‘what if’ scenarios. Typically the mobile app is one component of a multi-tiered system; therefore, restricted functionality is to be expected.
  • Emphasize performance - Typically the mobile worker will perform a small subset of functions repeatedly. If this is the case, those functions need to be processed quickly. Complex processing that can be deferred should be deferred as a post-processing activity. Mobile apps need to be small and efficient.
  • Use components - The use of software components, whether they are ActiveX or Java, can speed up development and will result in more stable applications that require less testing. Stable applications are also high on the priority list for the mobile worker. Essentially, a mobile app is seen by a worker as yet another tool or instrument and he or she does not have time to reboot the device, troubleshoot, etc. Such activities as initializing instruments and baselining should be automated and transparent to the user.
Figure 4 shows a prototype field client GIS for a workforce management system. Typically, these systems are complex and multitiered; however, the field client app is limited to only those functions, which the technician requires to do his job. The component shown here is the field GIS client. The technician can download a list of jobs assigned to him from the dispatch server. He can then use the GIS to route the jobs, display the jobs, as well as his position in a map view. He can then complete the ticket, and finally transmit completed tickets, as well as his position and status back to the dispatcher. All functions are available from the base GUI and toolbars are used over menus.


Figure 4. Work Order Mapper prototype


Figure 5. Task list of downloaded and assigned work orders

Conclusions
Mobile GIS will continue to expand into broader markets with ever more imaginative applications as business pressures continue to demand higher levels of interconnectivity and distributed applications. Clearly, the tapestry or ‘weave’ of the World Wide Web will become thicker and more tangled as faster, newer, more exotic devices are grafted to it either with or without wires. Clearly, the GIS/GPS combo is here to stay, as it becomes more ingrained in the mass market. Some of the changes just around the corner are: The integration of GPS chips within a sizable percentage of computers and other devices will turn the web into one giant personalized map, making demographic data instantly accessible. Such devices as two-way GPS pagers will become commonplace and give law enforcement much needed help in finding missing persons.

Viewed from another perspective, the trend towards higher levels of interconnectedness, the linking together of distant people and places, and the rich multimedia nature of the experience has allowed us not only to exchange ideas and art but it has also created the perception of the web as an evolving, self-organizing, living thing. The real marvel is not whether or not the network or the computing machinery will become intelligent but how the interaction between man and this new sensory web can be viewed as the next step in our evolution.

In 1987, C. Negroponte, the media wizard at the MIT Media Lab, commented on this trend by saying: “With flat panel technology you will find that every license plate, wine label and price tag is a display, our appliances will collectively have more processing power than our computers and your left cuff link will talk to your right cuff via satellite.”

References
  • ACSM/UCGIS, Research Priorities for Geographic Information Science: Cartography and Geographic Information Systems, vol. 23, no. 3, 20 p.
  • Brand, S., 1987, The Media Lab – Inventing the Future at MIT, 285 p
  • Dyson, G., B., 1997, Darwin among the Machines: The Evolution of Global Intelligence, 286 p.
  • Hurn, J., 1993, Differential GPS Explained, 55 p.
  • Motorola, 1995, Wireless Data Communications: An Overview, 16 p.
  • Motorola, 1995, Wireless Data Communications: The Choices, 27 p.
  • Olofson, C.W., Kusnetzky D, and Hwang D., 1997, Putting the Enterprise in the Palm of Your Hand: Occasionally Connected Database Technology from Sybase: An IDC White Paper, 98 p.
  • Sessions, R., 1998, COM and DCOM: Microsoft’s Vision for Distributed Objects, 492 p.
  • Shapiro, A. L., 1996, The Control Revolution: How the Internet is Putting Individuals in Charge and changing the World, 284 p
  • Wong, H., 1999, Jini and Network-Enabled Devices: Dr. Dobb’s Journal, #301, July 1999,pp 21-25.
  • Wood, D., 1992, The Power of Maps, 248p.
© GISdevelopment.net. All rights reserved.