In this article I will try to tell how the taxi market is developing and how the development of information technology influences its formation.

Retrospective

It was: The operator sits at the phone, accepts and records requests from customers in the notebook. Then on the radio interrogates the drivers for who will take the order. Many orders were not executed because such an organization of work has its limits on the number of machines simultaneously on the line.

It became: The phones turned into IP telephony gateways with recording of conversations, the notebook turned into an automated workplace and a server that determines the client’s coordinates at the address and appoints the nearest driver (or shows the order to the closest drivers and who will take it there). Accounting is automatic. Performance has increased. Now, it has 80-85% of all received calls for good service.

Will: People stop calling a taxi by phone. More precisely, they will stop calling. The call will be made by pressing one button in the mobile. Compare yourself: what is better, spend a minute trying to get through, explain where you are and then wait another 30 minutes or press a button, after 30 seconds find out what the car will be for 5 minutes and do nothing else? I'm not saying that if you travel around the country, then you do not really want to memorize different numbers.
')
“No, son, it's fantastic!” You say. And I will say that in some places it already works.

It is clear that for the imminent coming of the transport paradise, changes are needed in the current situation. So what needs to be changed?

"Button"

To call a taxi button, you must at least make a button. This is a client mobile application. The main task of the application: to save client time and work as simply as possible. You can make it very differently. To vskidku options are:

Phone book
It's not even a button, but just a set of phones by city. It’s better than looking for numbers yourself, but that’s where the advantages end. An example of the implementation he Hadnoidea.

There is also a western version of the Cab4me . They offer taxi drivers to highlight on the map the area they serve. When a client launches an application, it sees everyone who works in the area. Sometimes the app shows the nearest taxi ranks.

Button Application
Enter the phone number in the application (register). Press the button, the number and coordinates are sent to the taxi service. Then everything happens as it is now (they call you back, send SMS and so on). The disadvantage is that you are in the dark for a long time as to who will come to you and when it will come to you (remember about 85%?). I do not know living examples. Maybe you try, username ?

Application console
Register. Fill out a simple form. More precisely, one field "where" and click "order". Next, follow the status of the application. If you want, they will call you back to confirm the order or to inform you when the car will arrive. When ordering, you can select the class of the taxi (along with the fare) and specify other preferences.



Such applications with a confirmed number and details of the trip can be sent directly to several taxi stations or even to the exchange. Theoretically, the chance of execution of such an order increases. In practice, there is still a lot to be foreseen and done in the application, before the execution of orders reaches normal or increases it.

Our team made such an attempt and it turned out Teleport . Of course, we still have a lot of work to do: we need to connect more parks, add cities, reduce response time and car delivery time ... But! We already have loyal customers and a certain amount of orders daily.

There is also a variation on the topic, when, after placing an order, you are offered options to choose from: to go more expensive, but faster or cheaper, but wait. All decorated in the form of a list of proposals from different taxis.

Russian analogues:

• estaxi.ru (in development),
• taxionline.ru (works)
• InTaxi (in development).
• Yeehay is rather an exchange, but more on that below.

A separate item, nightmarish realizations from regional carriers:

• Java taxi console WELLING - I managed to make an order only on the second day of picking. Can be used as a test for intelligence.
• Similar for ease of use version of the taxi WEST (site down)
• WapTaxi is a Java application. There is no site, but there is a forum. They work for two years as.

There are other similar attempts. But some of them do not even run.

Western counterparts:

• TaxiPal - according to applications working in Europe and the United States. In fact, the scheme with ordering and tracking only works in some countries. In the rest - the application works like a phone book.
• GetTaxi.com - press the button, then follow the map as the car goes to you.
• Wheely - something related, still in development; makes a person fluent in 4 languages ​​in a team with a committer in Rubinius, and a cool objc developer, a former designer from a studio on Gazetny , the son of the Perm governor Oleg Chirkunov. Why, then, in the "Western counterparts" got? And this is the right question :-)

I think this list is far from complete. The idea of ​​an application to call a taxi excites many bright minds!

Web application
Why write for different platforms, when it is much cheaper to make a website for a mobile browser? Glory to Gd, they (browsers) support geolocation functions and even databases on the client. But, as you understand, not everything is so great with web applications. Apparently, therefore, there are no analogs that match the quality of “native” applications. Although, perhaps, just a couple of correct comrades have not yet been found ;-)

Increase% execution of orders and decrease response time

The second problem is performance enhancement. Formally, it is solved simply: you need a lot of cars. In fact, it is necessary to ensure the balance of drivers and orders in the system at any given time, and at every point in space.

It is clear that based on the law of large numbers, large parks have advantages. But even the largest parks in Moscow "do not take out" of all. I do not know for sure what the reason is: lack of machines or imperfection of technology. I think both have a place. The bottom line is that on the basis of these problems, there are attempts to solve them by a global union of all. But again, you can combine in different ways ...

Exchange
It's simple. There are exchange orders. Some give orders that can not fulfill; others take them and perform them. The commission, which gives the driver sawing about in equal proportions. Again, for reasons unknown to me, top taxis do not participate in the exchange exchange. Perhaps in order not to spoil your image in the eyes of customers. Well, or, for example, you manage a park where more than a hundred of their own cars. All cars are fresh Mercedes E-class. Price segment - premium. To give the order to your client? Yes, no one, in general. Do not send Logan there.

Another problem with exchanges is that they are intermediaries and add their own “noise” to the process of technical interaction. Those. in fact, the order needs to go through several gateways, each of which is owned by a separate company and each of which is built and operates according to its own principle. There is no single standard!

There are several exchanges in Moscow. In St. Petersburg, everything is average. In other cities, the exchange of exchange actually does not exist, or I know nothing about it. (Do not confuse the exchange with the control room, which ate regional tax stations and fully manage the flow of applications for each.)

By the way, the project Open-taxi , already mentioned on Habré, is also essentially an exchange, with an order interface for customers of their own production. An example of a successful exchange in Moscow with a good software: RBTaxi

Own network
So it turned out that the top taxis at the same time are the developers of automated workplaces for taxis (exception Infinity). The reasons are trivial: paranoia (ibid is our all!) And the fear that the base with a real list of applications may fall into the hands of the tax inspectorate (and this is the inevitable penalty for almost any park). Accordingly, if your software is used by 100 taxis in different cities, then it is a sin not to unite everyone into a common network and not exchange applications. Examples are the same: taxionline.ru and estaxi.ru

Again, there are no examples of the global dominance of specific software in a particular city. Accordingly, there can be no significant result in increasing the% performance.

In addition, if you have already been doing software for corporate clients for 5 years and are used to usability in such a large-scale project as an automated workplace, they are not the most important thing (well, that is, you think so), to produce a good client application you will have to do serious brain refactoring in themselves and their employees. The experience of the WEST taxi and taxi VESET (they are the Leader) described above clearly shows this.

Crew equipment, drivers

The way the taxi crew is equipped also affects the speed of processing the application. It's one thing if the car is equipped with a taximeter with a navigator, in which there is support for calculating the route with regard to traffic jams. And it is completely different if it is an old Nokia with a Java-application, where only two addresses “from where-to” and the phone number of the client are displayed. There is no need to talk about instant notification of the client about the status of the order and the position of the car.

In addition, different services have different policies: somewhere, the driver can take only the orders he likes, somewhere he is obliged to carry out all appointed. The problem is that in this case the interests of the driver conflict with the interests of the client.

Cost optimization

When there are many costs and imperfections in the business process scheme, you have to either increase prices or sacrifice quality. And this means the existence of a high minimum cost of the order, margins for travel from the airport and so on.

Who will win ???

Those who see at the root and successfully master the reading to this place will say: “I understood! You have to do everything yourself, in the form of a service! And the button, and the workstation and software \ device for the driver! "Thank you, cap. This gorgeous idea is trying to embody the guys from Taxofly . And also GetTaxi, about which I already wrote above.

But with such a monstrous idea, it is quite problematic to enter the formed market. Etozh need to find many drivers at once and many customers at once and start it all at the same time! Here the advantage is the existing developers of workstations. You just need to learn how to make client applications.