Authors: Igor Zubchenok, Stepan Churiukanov
It’s no secret that there are natural inaccuracies in computing object’s coordinates through GPS. When a receiver stays in a single location for a certain period of time, the system detects a set of different coordinates that lie within 5 to 100 meters from each other, instead of detecting the same coordinate. If this data is sent and saved in the database, it will cause redundancy and inaccuracy. For example, it may seem that the truck is moving, while in reality it stands still. As a result, the system will be winding mileage. We faced such a challenge when developing a route tracking module based on the GPS technology for one of the Logistics and Warehousing applications. We’d like to share the solution found to give you ideas on how to prevent major problems with GPS navigation on Pocket PC devices.
Too Many Records Means Too Many Traffic
Imagine that a truck is parked overnight and the driver does not turn off the Pocket PC device with GPS that sends its coordinates to the server. In case the truck stays in a parking lot for 10 hours and its position is calculated every second, the server will receive 36,000 records overnight! It will cause data redundancy, traffic overload, and total cost increase. If a record requires at least a dozen of kilobytes, irrelevant data sent over night can occupy hundreds of megabytes.
Solution #1: Narrowing-down Time and Distance
At first, we created two parameters that had to be considered for proper position calculation: time and distance. The system was made to ignore any position changes within 50 meters, and the inquiries were sent every 5 seconds. The values of these parameters can be configured, of course, to adjust the level of accuracy. If the difference between the positions was over 50 m in 5 seconds, the device sent records to the server. If it was less, it waited to send records till the required distance of 50 meters was passed. It did help compute coordinates with a given accuracy.
This solution worked out but we still needed more precise calculations of the distance passed by the vehicle. In the next post, we will describe another solution that improved GPS accuracy.