Use Case
Car sharing services are very common nowadays and are being operated in various cities around the world. In order to further enhance these services and the experience of a car sharing, experiments have to be made in a controlled environment. Mbots are used to simulate the behavior of a car, which is connected via various sensors and devices to the network. Our goal is to implement an intelligent priority queue system, that will optimize a resource use and contribute to the environmental protection.
We have three classes of users:
– Government – the mBot is used in an emergency, such as police chase or ambulance service. This type of user has priority over other classes of users.
– Premium – user with premium account, it has higher priority than freemium user, but lower than government user,
– Freemium – user with the lowest priority.
We also classified the routes. There is the red(short) route, that goes through the city, and green(long) route, which is out of city(highway), where the mBot can go faster. The mBot then makes assumptions about current traffic conditions and depending on these picks one of the routes.
The queue works by the ”first come, first serve”- principle, so the first user would get the car. However, if the priority class of the second customer is higher than the one of user one, the car will take user 2 to his preferred destination. Since, we have multiple priority classes different and interesting possible user queues can be instantiated. Consequentially, a rule base system was deployed to manage the queue. We also concluded that the classes ”premium” and ”freemium” are able to share rides respectively.
Experiment
We have documented three use cases. In first use case there is one government user, that goes to his preferred location. After that second user drives to volleyball court and third user is rejected, because his destination is different from preferred destination of mBot.
In second case we had two government users. mBot drives the first government user to emergency destination and than goes to “lockdown mode”(due to possible emergency situation in the city).
In third use case the government user also drives to emergency destination. Next in the queue are premium user and freemium user. Freemium user reaches his destination this time, which is same as a prefered destination of mBot.
Results
We successfully were able to build all the randomized user queues with different user priority classes and destinations.