GIS - GPS Integrated system for Call taxi owners - A case study in Besant nagar, South Chennai


Rama narayanan P.A.
arun_anna2000@yahoo.com

Raja Vijay Singh G.
rajavijaysingh@yahoo.co.in

Chidambara raj P.
pcraj_au@yahoo.co.in

Institute of Remote Sensing, Anna University

Abstract
The present paper deals with a GPS-GIS integrated system which when fully implemented helps the fleet owner of a Call Taxi company to continuously monitor his fleet of vehicles on the computer. In addition, he can take such decisions as the closest call taxi to a service call, shortest route to reach the service call and the quickest route in terms of time. The paper uses the Arc View software and its extensions to implement this system. In addition, Geotracker analyst module was used as an interface between the GPS and the GIS System.

It is hoped that the fleet managers would find this software useful to implement a real time vehicle location and optimum routing system. The results were generated in Arc View environment

The existing scenario of the call taxi network in Chennai is such that it relies at a human judgment before deciding the call taxi to be sent for various service calls and also regarding the route to be taken by the driver. The network manager at present doesn’t have the knowledge about the real time positions of his call taxies while taking decisions. Besides, the drivers of the individual call taxies are at liberty to ply the call taxies for their private gains without the knowledge of the owner. All these deficiencies exist in spite of the fact that Call Taxies have a great potential in Chennai city where the public seems to be disenchanted with the auto rickshaw drivers.

The present system aims to make the call taxi network to be more effective and efficient both to the owner of the fleet and to the customer. It will help the fleet owner by being more vigilant and save fuel costs and time. If implemented, this system will help the call taxi system to be more professional and modern in its approach.

Study Area Description
The Besant Nagar area in South Chennai is the study area used to design this system and used to test this system.

The area had the following geographic coordinates:

Longitude :	80° 15’ 55” E to 80° 16’ 13” E
Latitude :	12° 59’ 52” N to 13° 00’ 09” N

The Besant Nagar area was selected because it has lots of streets and inter connected roads thus offering good scope for network analysis. Although, Besant Nagar has been used for this study, this system could be extended for the whole of Chennai city.

Data Sources
This paper requires sufficient amount of spatial as well as non-spatial data to implement the integrated system. These are listed below:

Spatial Data

• Line Data
  Road map of Besant Nagar area at 1:2000 scale supplied by CMDS, Chennai.
  
  
• Point Data
  Call taxi locations provided by the GPS receiver. Control point coordinates for map preparation obtained by a GPS survey conducted in the study area.

Non Spatial Data
The following data was collected by conducting numerous ground surveys in the study area namely Besant Nagar in South Chennai.

• Street names and addresses
• Corporation speed data for various streets
• Street conditions
• Important land marks

Map Digitization and Paperion

Required Map Accuracy
The digital map forms the heart of this real time automatic vehicle location system. The end user is going to monitor the position and the path of his vehicle on this map. The map has to be user friendly and should give important landmarks around the area so that the user knows the ground location of his vehicle. The accuracy of the map is around 3 meters.

The positional accuracy of this map is also a very important factor to consider. This is because of two important reasons. They are:

• The whole objective of the paper is to display the position of the vehicle and hence if the map is not accurate enough, the high precision GPS coordinates would not be of any use. Also, since the coordinates of any vehicle are going to be changing quickly the map should display these changes effectively.
• The communication link of the system is going to introduce certain errors because accurate systems require high costs and infrastructure. Hence, the map needs to be accurate enough to offset the uncertainty in position caused by weak signals.

Considering these factors in mind, it was decided to go for very accurate maps of the study area. The DDP map sheets of CMDA for the Besant Nagar area gave the required accuracy as well as the desired scale that would be suitable for this paper.

Map Scanning and Vectorization
The maps given by the CMDA were A0 size and the first step towards digital map creation is the scanning of the map products. Once the scanning was done, the file was stored in AutoCAD .pcx format.

Next, the Vectorization of the scanned map was done by using the onscreen digitization option present in AutoCAD. Once the whole map was digitized and saved as .dxf format, it was imported to Arc Info software as coverage. Using the edit features of Arc Info, the various node errors were removed and labels were added to denote various roads and streets.

Identification of control points and control survey using GPS
Once the vector form of the map was ready, the next step was providing planimetric control to the map. The main requirements for this purpose were well-located and accurately surveyed control points.

A reconnaissance survey of the Besant Nagar area was conducted to locate suitable control points, which could be easily located on the map as well. About sixteen control points were chosen which could be easily located on the map.

Control survey was done using GPS to obtain a high degree of accuracy. Initially, a base station was chosen in Besant Nagar area and survey was carried out in relative positioning mode. This was done by keeping a receiver at the base station and simultaneously taking readings at the various control points using the other receiver. The radial line method was used to compute the coordinates of the various control points in the WGS84 system. Out of the sixteen points, readings were taken for twelve points.

Geocoding of the map to the WGS84 system using PC Arc-Info
The various control points for which survey was carried out were identified on the map and they were entered as tic points in the Arc Info coverage. Tic points are used for transformation of the coverage from the digitized map coordinate system to the WGS84 system. This is necessary because the signals transmitted by the vehicle will contain its coordinates in the WGS84 system.

Tic ids were given for just seven of the twelve control points so that five points could be used to check the accuracy of the transformed map in the WGS84 system. The transform command of Arc Info was used to transform the whole coverage into the WGS84 system.

To check the transformation and the accuracy of the geocoded map, the where command of Arc Info was used to display the coordinates of various points selected randomly across the coverage. In addition, the map coordinates of the five check points were also found.

Using Geotracker and Network Analyst

Geotracker Module
The Geotracker is an extension of the Arc View software that acts as the interface between the GPS and the GIS by displaying the positions of the call taxies in the form of latitude and longitude on the map.

Network Analyst Module
Once the vehicle can be tracked on the map, the manager has to be able to use the Network Analyst module to find the closest call taxi to a particular service call and to find the optimum route from that call taxi location to the place of service call. Arc View offers a separate extension called Network Analyst that helps the manager to take these decisions. Before using this module, a suitable line theme is required on which the network analysis could be carried out. Also, the call service locations and the real time call taxi locations could be loaded on to the viewer. If the analysis is to be carried out in the off line mode, then the GPS file containing the locations could be loaded into the viewer.

The basic output of this paper is a graphic display of the real time call taxi locations, closest call taxi to service call, the optimum route and the directions to navigate through the optimum route. All these options are best viewed on a computer screen. The results are not exhaustive by any means and are just representative of the numerous outputs and options available from this system.

Methodology


Results and Discussions
The transformed map gave excellent results as far as the accuracy was concerned. This was verified by checking the map coordinates of the five checkpoints with the coordinates obtained from the control survey. The difference was less than a tenth of a second that converts to a very good positional accuracy if we assume that one second corresponds to about 30 meters on ground.

The group has aimed to get as accurate a map as possible within the available time. Specific attention was paid to the control survey so that the control points were of good planimetric accuracy. It is hoped that this accuracy is good enough to achieve the end result of this paper.

The next step was the preparation of the Geocoding index for the theme containing the streets. The Geocoding index tells the computer the address format to be used while identifying various point locations on the map. The address format chosen to represent the addresses was the US Single Street numbering system without zone information. Before doing this the fields like roadname, fr_add, to_add, distance, time and speed were added in the theme’s attribute table. Once the required fields were entered into theme’s attribute table, the theme was ready for the Geocoding index to be built. To do this the following option in Arc View main menu was followed:
Theme ® Properties ® Geocoding editor
Using the above option, the Geocoding index was prepared. The various address locations collected from the ground survey were entered into the respective fields and using the address locator, the addresses were displayed on the map. Next, certain fields were added to the theme’s attribute tables that are required to determine the optimum route to be determined between the call taxi and the service call. Thus this proves to be a better system for Call taxi owners in Chennai instead of present conventional system.