Travel time estimation based on previous experience - Pre-study and prototyping

Travel time depends on various factors, which can be described by data coming from sensors. The author makes an assumption that for the same trip conditions travel time will be the same, and if we can collect enough information on the current trip conditions and find a matching trip which took place in the past, we can estimate the travel time for future trips. The project aim is to design and prototype a system capable of collecting this data, organizing, storing and using it to find matching trips, with the real-time performance being the main consideration. The scope of the system is limited by the needs of a logistic company which wants to be able to track its vehicles and estimate their travel times. The resulting system is tested in various settings to find out how well it performs. The author identifies the settings which are suitable for the particular implementation and suggests further improvements which are meant to extend the settings