Visualising trips and travel characteristics from GPS data

Road & Transport Research,  Jun 2003  by Stopher, Peter R,  Bullock, Philip,  Jiang, Qingjian

• E-mail
• Print
• Link

GPS devices

The passive in-vehicle GPS device, or 'Geologger', currently in use at ITS is shown in Figure 1. The device is activated whenever the ignition of a vehicle is on, and operates independently of the driver. It is worth noting that ITS has been undertaking some experiments with a 'wearable' GPS device, which can be carried by respondents using public transport. This device is still under development, but the invehicle devices have been used extensively, and are the primary focus of this paper. The main issue to be addressed is how to present the information contained in the records of these devices so that it is understandable both to the survey respondents and to decision makers. There is also an issue of how to present to analysts and decision makers additional information that is present in the records, but that may not be needed in interacting with respondents.

An example of the type of data available is shown in Table 1. The data available from these devices is the standard output data from a GPS antenna, consisting of:

* Latitude and longitude in degrees and decimal degrees, with hemispheric (E, W, N,S) designation

* Altitude in metres above sea level

* Heading in degrees from north

* Coordinated Universal Time (UTC Time, or Greenwich Mean Time)

* Coordinated Universal Date (UTC Date)

* Speed in km/h

* Horizontal dispersion of precision (HDOP)

* Satellites in view.

DATA MANIPULATION

Simple computations

Certain items of the data require manipulation to make them useful and usable for Geographic Information System (GIS) representation. First, the record must be re-ordered so that it can be recognised by the GIS software as a geographic file. The software being used for this is Caliper Corporation's TransCAD(R) software. This requires that each line of the file is uniquely numbered with an ID number, and that the first column after the ID is the longitude, and the second is the latitude. Further, longitude and latitude are to be expressed in millionths of a degree, with negative values indicating south or west, and positive values indicating north or east. The compass quadrant is therefore removed, and replaced by the appropriate positive or negative sign for the latitude and longitude. Next, the UTC date and time are converted to local date and time. At the same time, these records are converted so that they are recognised by the database procedures to be times and dates, so that calculations involving times and dates can be performed.

Some additional computations are needed to enable better visualisation of the results. First, the elapsed time and distance between successive track points are desired. These are used to determine trip lengths in time and distance in subsequent steps. For the purposes of better communication, it is also desirable to add the day of the week to the date and time. Additionally, acceleration or deceleration between successive points may be desired, and are calculated from the speed data.

Detecting trip ends

The most challenging part of the data manipulation is breaking the record into individual trips. To do this, it is first necessary to understand how the data are recorded. For the in-vehicle GeoLogger, we assume that the car's ignition turns the accessory slot on and off. In this case, the GeoLogger is on only when the engine is running. (There are some cars in which the accessory slot is always live. We have excluded vehicles of this type from our surveys, so that we are able to determine times when the engine is off.) For this situation, long elapsed times between two successive points would clearly indicate the end of one trip and the beginning of the next. A situation of this type is shown in Table 2, which also shows some of the results of data manipulation.