Team #09 // EasyTicket

Executive Summary

Our solution EasyTicket is a novel method for handling public transportation tickets in an agile and seamless way without the need of any interaction by the user. The solution is based on TAPAS hyper-precision geo-localization system and contains a real-time monitoring system of public transportation, to enable optimization for operators. The users of our service – by just keeping their phones in their pocket – can be detected in case of using any transportation services, and directly charged with the exact price of the journey, without any difficulties with buying tickets and check-in and out. We collect precise, valuable data that can be analyzed and used to optimize public transportation, while also keeping privacy and security in mind. The incoming data allows the operator to track every act happening in the transportation environment in real time, enabling immediate reaction to failures and accidents. The operator is also able to balance its resources according to the present load and conditions, therefore saving costs and reducing pollution.

Problem

Buying and managing transportation tickets can be a real hustle, especially in foreign environments. Having plastic cards and paper-based tickets and check in/out with them is inconvenient. Furthermore, operators have inaccurate and non-verbose information about usage of routes and unable to monitor utilization of infrastructure in real time.

Target user / Customer

• Passengers: they can use an extremely light-weight, easy ticketing system, where everything (from buying ticket to check out) is handled by the mobile application and infrastructure, without any direct interactions by them.
• Public transportation operators: they can utilize the collected data to optimize routes, monitor their system and their users in real-time, so they can react immediately.
• 3rd parties: data can be offered to ad companies to inform them where to locate ads on vehicles and in the urban areas. We can also include vehicle manufacturers to provide information on space usage in vehicles, to develop interior designing. Urban planners are also potential clients.

Your solution and how the concept is feasible

Our solution is using TAPAS technology to identify journeys of users in public transportation, while excluding the need of the usage of any hardware infrastructure for buying tickets and check-in/out. The infrastructure of the system is really light-weight since it uses passengers’ mobile phones (which run our application) as a sensor system, to provide data about the transportation environment. Using the 1-2 cm accuracy of TAPAS for localization is crucial in our application, since we have to be 100% sure that we identify correctly every journey. The capability of charging the users correctly means the trustworthiness of the service.

There can occur two difficulties, one is battery consumption and the other is signal loss. In order to optimize battery usage we would introduce a rest mode, by reducing the frequency of sampling based on the user’s behavior, location, geo-location history. To solve the issue related to signal loss, we would also use the acceleration sensor of the phone and external base-stations underground to make sure about accurate detection. Data outage, caused by signal loss would be refilled with the help of a smoothing algorithm, using edge-computing.

Data privacy is a key aspect in the design of our service and business, therefore we emphasis to properly fulfill any requirements regarding it. We eliminate the possibility of connecting any individual with any information given to 3rd parties. We solve this by introducing anonymized tracking ids and aggregating information, furthermore reducing the location information to parts, which are strictly transportation related.

Four unique value propositions

1. Seamless usage of public transportation without using physical tickets or cards for check in/out
2. Accurate and optimized routing of public transportation, real-time monitoring
3. Cheap and light-weight infrastructure
4. Versatile and precise data collection for urban planning and other 3rd party clients

Architecture

Our design of architecture is purely built on open-source tools and also highly scalable – ready to handle millions of users – which is an important requirement in case deploying it in big city environments. We use an extra layer of load balancers to deal with heavy load and also providing a proxy to ensure security. The application runs in a dockerized environment, orchestrated by Kubernetes. For the ticketing service, we use Postgresql relational database for transactional queries. For the analytical side of the service, we choose Kafka as a messaging queue to be able to process data in real-time. It directly connects to a Spark Streaming cluster to parallelly process the data of millions of users. The transformed data is then loaded into our HDFS data lake. From the data lake we transfer the data into two data store units, one is a data warehouse for analytical queries the other one is a Neo4j graph database for queries related to connections between entities of the transportation environment. We build up the Data service API from the same building blocks as the Ticketing API. The application is implemented in NodeJS with ExpressJS.