Distance Calculation In Case of GPS Loss (Billing)

We depend on the location data sent by the cab to track it in real time and find out the distance traveled/route followed once the trip is complete. In order to get the location from the cab, we rely on the GPS location received by the cab.

How Cab Location is Sent

The location in cab is identified by the driver mobile app sitting inside the cab. The driver mobile app switches on the GPS receiver of the mobile phone. In order to get the GPS location, the driver app needs to receive signals from at-least 4 GPS satellites in an unobstructed line of sight. The locations are received in form of geocordinates (latitude and longitude). The device determines if the geocordinates received are correct. If it finds them correct, it collects points for 10 meters distance and sends them to the ETS server. ETS server stores this data in its DB. So, this is how a user can see the cab's location in real time or after the trip is complete on a map.

Once a trip is complete, ETS also calculates the distance traveled by the cab based on these locations. However, we may sometimes have a situation where GPS data could not be obtained for a period of time. This may happen if device is unable to connect to at-least 4 satellites or is in an area with tall buildings where device does not have a clear line of sight to any GPS satellite. The possibility of such an event, on average is less than 0.50% on a Moto E device. 

But, suppose it happens in a trip and now after the trip, when a user is trying to create the bill for the trip, the user has no idea what is the distance traveled or route followed by the cab when the GPS was lost. The user is thus required to take a guess and fill in this trip distance manually.

To fix this, we have developed a predictive model which tries to identify if a trip had a GPS loss and then tries to predict the route and distance traveled by a cab. The solution has two parts - 

1. A stand-alone program which can be executed for any date range to back-fill the GPS data lost during that time.
2. At the end of every trip, if the GPS back-fill property is enabled, ETS will automatically back-fill any location milestones which are missing in the trip

The solution tries to re-create the path followed by the cab when GPS was lost by determining the time when an employee signed-in or signed-off a trip. It also looks for employees who were No Show on the trip and had not cancelled the trip. It then determines if the location was known at these timestamps - if not, it tries to determine the sign-in/sign-off locations based on employee's schedule or default locations. Based on this information, ETS re-orders the milestones and calls Google APIs wherever it observes missing information between any 2 nodes. 

The modified distance is summed into the original trip distance reported on basis of GPS data. On the ETS, user will see the route and distance on the map for the trip as predicted by the back-fill program. A comment field will also be visible in the line against the trip where user can see what the modifications have been by hovering the mouse over the comment icon.

The solution will help ease the hours spent by the transport team during billing in guessing the trip distances and updating the same. It will also help mitigate the risk of manual errors and create transparency in the system by having a single system of route prediction. 

The current solution works for the single office set-ups. The other limitation of the solution is that it cannot work for guest users (non-employee) in trips if their location is unknown when they boarded/dropped off the cab.

We are committed to continuously work on this solution to improve it further. Please contact am@moveinsync.com to enable this solution on your premises.