|
Building POI LBS and its Indian Market Potential  Vinit Kapoor Sr. Technical Lead Persistent Systems Ltd. vinit_kapoor@persistent.co.in  Sanjay Khuntia Software Engineer Persistent Systems Ltd. sanjay_khuntia@persistent.co.in Abstract Location Based Services (LBS) have become popular and a potential revenue generating source. They have already proved their potential in developed countries and show a promising future in countries like India. This paper presents a technical and business study of simple Point of Interest (POI) LBS from Indian market’s perspective. It analyzes different architectures for deploying this service, and their technical and business feasibility. It further goes into details and discusses a case study of a successful exercise undertaken at Persistent Systems Limited (PSL) of building a POI LBS for GSM/GPRS based network service. Later on it briefly analyzes the business potential of POI LBS for Indian market. This overall study of POI LBS can help vendors, researchers, and developers to use these concepts and information in making further progress of LBS in Indian market. Introduction Location Based Services are becoming one of the popular services that can be offered to the mobile subscribers. These are services which take into consideration target’s location for adding value to the service. If we go by literature, these falls under context aware services where the context considered here is the location of the service seeker. There are various kinds of services that can be provided to mobile subscribers based on their location information. These include but are not limited to emergency services, location based information, navigation, tracking (fleet, logistics), location based advertising, entertainment and social networking. More details about them can be found in. In mobile domain, LBS fall under the class of Value Added Services (VAS). LBS have already proved its potential in developed countries like U.S.A, Japan, Korea. However, it is still an emerging field in markets like India where till recently focus of mobile operators has been on increasing mobile subscriber base. Considering this, the paper will be mainly focusing on basic POI LBS. The reminder of this paper is organized as follows: the next section describes typical LBS value chain. Then we cover details about building the POI LBS service, describing various elements of value chain, their expected functionality and various possible business scenarios depending on location technology being used. Following this we cover a small experiment performed at PSL about building POI LBS and then high level Indian market potential and feasibility check for POI LBS. Finally, we conclude with summary of the discussion. Typical LBS Value Chain To understand any service market, it is required that we should go through the value chain (also called supply chain). This depicts the various entities involved in making the service available to end user. Figure 1 depicts a typical value chain for LBS. Such value chain dynamics have been described in literature. We will cover it briefly and discuss it in more detail from the point of POI services.  Figure 1: Typical LBS Value Chain Various Entities Involved, their Role and Required Interfaces are as Follows – • Mobile handset (also called target): Mobile handset whose location is to be determined. This could also be LBS user for certain kind of services like POI service. For services like friend finder, LBS user and target are different. • Network Infrastructure: This represents various wireless network elements like base station, MSC etc. • Position Determining Entity (PDE): These are special network elements required in wireless network to determine location of mobile handsets. They interact with other core network elements of wireless network. • Location Gateway (LG): Acts as an interface to outside world to accept request for location determination. It receives request for location determination of mobile devices from Location service provider. To determine location it interacts with PDE and other wireless network elements. • Location Service Provider (LSP): This provides the basic required infrastructure for implementing various location services like POI, navigation etc. It usually handles spatial data management, indexing of data and the aggregation of data from different content providers etc. • Content Provider: This element is responsible for providing the actual content which is to be delivered to the LBS user. The content categories include points of interest data, navigation data etc. • LBS User: This is the end user who normally initiates the location service. All the major interfaces between elements from target handset till LSP have been standardized by various organizations like OMA, 3GPP, 3GPP2. This helps in addressing the interoperability between elements. The value chain depicted above also depends upon the location determining technology being used. This value chain mainly addresses the technologies which require network involvement in determining the location of handset. In the following section, we consider the impact of location technology on the value chain. Point of Interest LBS As described in above sections, information search considering location context is one of the basic LBS that can be provided to customers. However, though simple, there could be very different needs of the users from this service. We describe below some of the different possible usage scenarios. Immediate Usage: This includes situations when a user wants to utilize the search information innear immediate future. For e.g., consider the case where I went to market to buy a book but remembered that I should also check on some other item like shoes. Here I would like to utilize the search information immediately before returning to home. Deferred Usage: This includes situations where search information would be utilized late in future or in some planned manner. For e.g., consider the case where I want to buy certain things but I am planning to go for shopping on weekend. Here I would collect say all the business listings falling within 4 km of radius of present location and optimally plan my shopping. Piggy Backing: This considers the case that I am travelling from location x to y and also want to check if I can get certain item which I wanted to buy along the way. An example of this is, I am travelling from office to home and want to buy a book along the way. Here buying book could be of a second priority hence I do not want to deviate from my travel path though I may be ok to take any other alternative path say the one which goes through a market. The above scenarios form a base around which the information search service could be developed. Further, the first and second scenarios call for considering the temporal aspect during search as well. This is described below in further detail. Similarly, third scenario calls for integration with navigation related GIS information. To understand temporal context further we consider following classification of information – Static Information: This category includes information which doesn’t change often. As an example, a shop keeper typically keeps his shop closed on any one day of the week. Or there are certain markets which are closed on some particular day of a week. To add further, there are tourist places which are closed on a particular day of a week. Dynamic Information: This includes current presence information of any entity. This is very similar to the presence information of buddies in IM. As an example, certain shop owner closes shop say 1 hour before its regular scheduled closing time (static information). In the above three temporal usage scenarios considered, deferred usage falls under static information and immediate usage under dynamic information category. POI LBS Value Chain POI LBS value chain is similar to the typical LBS value chain. However, one difference here would be that LBS user would be same as the mobile target handset whose location is to be determined. Following is a description of some of the important elements present in the POI LBS value chain along with their expected functionality. LSP: This is main element which provides the POI service. Its required functionality and interfaces to be supported are as follows: Functionalities: A lot of knowledge related to this can be obtained from the internet search and indexing techniques. These need to take into account spatial and temporal dimensions. The kind of queries that need to be addressed are –
Interfaces: Content Provider: This includes content providers for business listings, navigation data and presence information (dynamic information). Local search listing and navigation data: Working mechanisms of these vary from implementation to implementation. How this information is transferred to LSP also depends on specific implementation. For these elements, providing regular updates of data to the LSP would be critical as accuracy of search results depends on correctness of data. No real time updates are expected on data as most of it should be fairly constant over a period of time. Hence batch processing mechanism can also work in this scenario. Location Technology Impact on POI LBS Value Chain There are various ways of determining the location of cellular mobile devices. This topic has been researched a lot. We describe here the impact of this on the POI LBS value chain. These technologies can be classified into following categories depending on whether the LSP gets the location information of user along with the search request or it needs to determine the location of the user using location gateway element. Ready Location Scenario: Here LSP gets location information along with search request (technology examples – GPS). This calls for direct interaction between user’s handset and the LSP. Technologies like WAP, web service can be used to implement this interface. Figure 2 shows value chain addressing this class of scenarios.  Figure 2: POI LBS Value Chain for Ready Location Scenario Non-ready Location Scenario: In this case, LSP needs to get the location of user through location gateway element (technology examples – AOA, T-DOA). For this LSP should be able to determine the appropriate location gateway to query for the given user. This could vary from carrier to carrier. To address this issue there are location aggregator (LA) elements which can be seen evolving in market. Figure 3 shows the value chain with LA elements in place.  Figure 3: POI LBS Value chain with Location Aggregator The table (Table 1) lists out various advantages and disadvantages of these scenarios from LSP’s viewpoint.  Table 1: Various Advantages and Disadvantages of these Scenarios from LSP Viewpoint Technical Feasibility: This section checks on the technical feasibility (including practical aspects) of POI LBS. As per our findings, most of the functionality described above is technically feasible to put in place. However, presence information functionality needs to be checked from the practical aspects. It needs to be determined as to how much would be the acceptance of this functionality among the users (business listing owners etc.). Implementation of this in user friendly way will also play an important role in its acceptance. Privacy Aspect in POI LBS: Privacy aspect for this kind of service should not be of concern since the user is the only one who would be giving its location information and using it. One way of handling this could be through privacy agreements. Experiment at PSL This section describes implementation of an experimental basic local search application which was based on CellID location technology. This experiment was done to demonstrate how different elements of LBS value chain interact to provide the service. There is a similar related study done earlier, however difference here is that we tried to demonstrate the complete LBS value chain by integrating with local search and content provider which is different than what was done earlier. The system architecture of the design is shown in Figure 4. This implementation uses the local web search facility provided by Google. Google’s facility requires the user to enter the location (text description of location like “Nagpur”) and the string to be searched for. Briefly, the way this service operates is as follows –
This service was implemented using J2ME, Mobinfo library for Symbian OS based on S60 2nd Edition Feature Pack 2.  Figure 4: System Architecture for Cell ID based POI LBS Application Limitations of this Approach: Accuracy of CellID method for location determination is not good (ranges from 100m to 10 km – range of a cell). Secondly, mapping of CellID to location string will require assistance from cellular operators to get this database created quickly. The other methods (community based) as indicated in could require long time to build this database to make it viable for service to be launched. Indian Market Potential In this section we will check on the different business models followed in Indian market, business feasibility of various models and finally the indications of revenue potential that we see for POI LBS in Indian market. POI LBS Business Models and their Presence in Indian Market In Table 2, we check on the revenue sharing aspects of different POI LBS value chain scenarios discussed earlier from LSP’s viewpoint. This is in light with the study that in Indian VAS market revenue sharing is skewed (benefiting operators more than other elements).  Table 2: Revenue sharing aspects of different POI LBS POI LBS Business Feasibility Check In Table 3, we check (by way of examples) the presence in Indian market of different elements of POI LBS value chain. This can help us identify gaps if any in deploying this service. Value chain element Indian players Global players  Table 3: The Presence in Indian Market of Different Elements of POI LBS Value Chain As seen from the above table, mostly all the elements of value chain are available (either through Indian or global players). Hence once the basic location determination infrastructure is made available by operators then market can expect the basic LBS like POI be made available to the mobile subscribers. Revenue Potential Indicators As seen earlier, there are many content providers available like local business listings etc for Indian market. Further, as per reports, the local search sites are getting good response. JustDial has recorded more than 0.1 million hits in a day] which indicates the demand for this service in Indian market. Adding location dimension should make the service more accurate and useful. Further, market research report indicates good potential for mobile search kind of services in India. It is predicted that VAS revenue share in telecom operator’s revenue will increase to around 10 to 20% by 2010 [It was around 7% by end of 2007. This indicates favorable environment for the POI LBS to show its potential if certain areas are addressed like large presence of low end mobile phone devices in Indian market. There are also indications in report that advertising based revenue models could help in making VAS services popular mainly by reducing the end user cost. This calls for integrating location based advertising (LBA) with location services like local search. Conclusions Following conclusions can be drawn from this discussion: References
|