This paper focuses on a navigation service (GreenGPS) for vehicles in which fuel efficient routs are computed. GreenGPS helps to reduces CO2 emissions, which has a positive effect on the environment. The idea behind this approach is participatory sensing, where the data is collected through different experiments and analyzed.

Question answer session

Different issues were discussed after the presentation.

a) How we can distinguish fuel efficient route from shortest and fastest routes?

Shortest route is determined by estimating the length of the route. In the fastest route, we consider the speed limitation, stop signs, and traffic congestion. In the same way, static (stop signs, traffic lights, distances travelled, and slop), dynamic (average speed), and car specific (weight and frontal area) parameters are considered in the fuel efficient route. As a result, if these parameters are considered, we’ll be able to find a route to save fuel.

b) Did authors consider different car models for experiment? What kind of parameters did they consider?

Experiments were conducted through different types of cars, and only two paramaters (car weight and car frontal area) were used for each car. However, they should consider other parameters (e.g., power consumption by electric modules inside the car, engine characteristics, brakes, etc.).

c) What is OBD?

OBD-II PIDs (On-board diagnostics Parameter IDs) are codes used to request data from a vehicle. Usually, vehicles are required to support OBD-II diagnostics, using a standardized data link connector. In this paper, authors used OBD-II to get data from the vehicle.

d) What is model clustering? Why did authors use model clustering?

Clustering is the task of assigning a set of objects into groups so that the objects in the same cluster are more similar to each other than to those in other clusters. Clustering is a main task of explorative data mining, and a common technique for statistical data analysis used in many fields.

Authors used only two parameters (car weight and frontal area) for each car that are not enough to analyze completely. Therefore, they tried to explore some results by considering car making company, car model, and car year. They estimated cumulative error for each car using model clustering.