I want to share my impression of the results of our company’s work with the distance matrix (DM).
I work in the delivery service of one of the restaurants, which belongs to a single chain of restaurants open in several cities in our country. One of the services of this restaurant is the delivery of orders throughout the city. The number of orders during the day exceeds several hundred. And, of course, the ability of the restaurant’s delivery service to quickly and efficiently ensure the delivery of the orders affects the satisfaction of our customers.
Our service is well equipped technically and informationally. Our software allows us to analyze the information that we receive in the form of geo-data from Google, as well as the distance api service.
So, our task is to develop a fast way to deliver the order by courier with minimal costs for the restaurant. To do this, we need to know thoroughly and analyze not one, but several key parameters. This is the distance between the given objects and various options for overcoming this distance in the presence of intermediate locations, and the time needed to spend on this overcoming.
At the same time, sometimes it is necessary to take into account the logistical nuances of the selected routes, in consideration of the manner of movement, because, depending on the distance to the customers, the courier can move both on foot and on a bicycle, motorcycle, or car.
And here a simple geographical map of the route will obviously not be enough. So, without a doubt, you should turn to DM API services to immerse yourself in information about all the nuances of a planned movement.
In general, there is nothing unusual in DM API. This instrument has been used for a long time for various purposes and in various formats. Now IT technologies have supplemented it by including the distance matrix in the api interface. And now the distance matrix has acquired the status of DM API.
It turned out that this approach significantly increased the possibilities of route planning for our couriers, because the api program, as a link-mediator, allows you to combine various programs and direct their functionality to implement the tasks in planning movements.
What does this mean in a practical case? The chain of actions is simple and convenient and consists of the following steps: sending the needful request and receiving the appropriate response. At the same time, the api interface uses its functionality to ensure interaction between the client application that generates the request and the server application that generates the response.
Since in our review we are talking about the development of routes, developers should use the api as an arsenal that provides the involvement of functionalities that allow them to calculate distance and time options between given locations (two or more). In this case, various nuances of the routes (traffic jams, regional rules, etc.), as well as the chosen mode of transportation (pedestrian, car, bicycle, etc.) can be taken into account.
In order to get the correct answer, first of all, it is necessary to issue a correct request. This means that the request must contain the required parameters set.
First, the request must contain an indication of the source location (one or more). Here it is allowed to indicate the location, both in the system of geographical coordinates and in the form of a postal address.
Second, the request must contain an indication of the target location (one or more). Just as in the previous position, the location can be indicated both in the system of geographical coordinates and in the form of a postal address.
Third, the request must contain a special identifier – the API access key.
Obviously, the more initial information the request contains, the more complete and reliable the answer will be. Therefore, if possible, additional information can be included in the request. This additional information in more detail you can be found in the guide for developers.
Our customers order food in different ways: by a phone call, by going to the restaurant’s website or by using a mobile application. This is where the forward and reverse geocoding service helps us. At the same time, the operator in the call center simply needs to fix the declared address. Further – the program of a DM api works. First, the address is converted into geo-coordinates, and when the program responds, the opposite is true – the geo-coordinates are transformed into postal addresses. In this case, the program calculates the distance and time directly from the restaurant kitchen to the customer.
The DM api response type is a JSON format already known to users. And, as we said earlier, the table with the results of route calculations can include several options for the requested route which can cover any points on the world map, not forgetting to take into account the various ways of moving the object along the route (on foot, by car, by bicycle, etc.).
The response must take into account the correctness of the request. There may be situations when the request was made incorrectly or the source and/or target locations were not found by the geo-program. In such cases, the DM api response will still be formed, but instead of information, it will contain empty arrays, and the distance matrix itself will contain the corresponding error indicators.
If the request is correctly executed and the necessary initial information is available, the response of DM api will contain the “OK” indicator.
Previously, we used Google’s API, but then we abandoned this service in favor of a more economical, fast, and convenient service – DM api. The process of changing services itself was not difficult due to the similarity of syntaxes.
Against this background, it is pleasant to note that in DM api there are all the necessary functions and tools that clearly and quickly respond to the diverse and sometimes unique tasks that its users set for it.
And we appreciated DM api completely and use it with great pleasure because after its implementation, our costs decreased by 4 times.