TY - JOUR TI - Capturing and analyzing human driving behavior to improve road travel experience DO - https://doi.org/doi:10.7282/T3BV7KRC PY - 2017 AB - People commute on daily basis and spend a considerable amount of time on road travels. Unfortunately, due to the dense population and concentration of economic activities in metropolitan areas, urban transportation suffers a lot of challenges, e.g. traffic congestion and parking space storage, which also affect each individual's travel experience. Previous studies have looked into the transportation condition sensing solution using various static or mobile sensors and traffic related recommendation services based on the sensing results, e.g. vehicle routing and parking recommendation. However, human behaviors as the most direct reflection as well as the final modifier of the traffic condition, have not been fully explored to close the sensing-and-control loop in the urban transportation system. This dissertation aims at improving drivers' road travel experience by taking human driver's behavior in vehicle routing and parking search processes into consideration. In the first part of the dissertation, we focus on the traffic condition sensing and routing service design, altering the isolated traffic sensing and the greedy routing strategy based on sensed traffic condition. Instead, we present a participatory system, called Themis, to consider traffic sensing and route planning as a whole: (i) By analyzing time-stamped position reports and route decisions collected from the Themis mobile app, the Themis server estimates both the current traffic rhythm and the future traffic distribution. (ii) The routing requests are then combined with the sensed traffic condition to provide route recommendations that minimize the travel cost of all drivers to proactively alleviate traffic congestions. Themis has been implemented and its performance has been evaluated in both simulation experiments using real data from over 26,000 taxis and a field study. Results from both experiments demonstrate that Themis reduces traffic congestion and average travel time at various traffic densities and system penetration rates. In the second part of the dissertation, we look into the crucial part of parking recommendation, i.e., fine-grained parking availability crowdsourcing and propose a solution based on human's parking and ignoring decisions: a parking decision immediately take an available spot while an ignored spot along a driver's parking search trajectory is likely to be already taken. However, complications caused by drivers' preferences, e.g. ignoring the spots too far from the driver's destination, have to be addressed based on human parking decisions. We build a model based on a dataset of more than 55,000 real parking decisions to predict the probability that a driver would take a spot, assuming the spot is available.Then, we present a crowdsourcing system, called ParkScan, which leverages the learned parking decision model in collaboration with the hidden Markov model to estimate background parking spot availability. ParkScan has been evaluated using real-world data from both off-street scenarios (i.e., two public parking lots) and an on-street parking scenario (i.e., 35 urban blocks in Seattle). Both of the experiments show that ParkScan reduces the error of spot-level parking availability estimation compared to the state-of-art solutions. In particular, ParkScan cuts down over 15% of the estimation errors for the spots along parking search trajectory even if there is a single participant driver. KW - Computer Science KW - Traffic flow KW - Automobile drivers--Behavior LA - eng ER -