Traffic Data Analysis

The traffic data analysis team aims to “make movement of people and goods more efficiently” by analyzing and utilizing various traffic data.

We conduct research to calculate an optimal travel route to avoid rain from rainfall data, and delay prediction from bus travel record data.


Real-Time Bus Arrival Time Prediction

In Aichi Prefecture, there are no bus arrival time presentation in bus stop sections. In this research, we propose statistical analysis on the bus operation history data collected by bus company to estimate arrival time in real time.

In addition, by providing the estimated arrival time and its accuracy, it enables users to use the bus more comfortably by increasing the user's choices of movements. This research also aims at development of the bus industry by mining knowledge and value from the travel operation data which has not been used effectively until now.


Bus Operation data Visualization and Simulation

By analyzing the bus operation data, we aim to discover and solve problems from bus operation data.
We develop a library that displays information of bus movement and passengers on the map by visualization library using deck.gl. In addition, we use this library to reproduce the actual bus operation on the map and discover new problems.


In addition, we develop a simulation environment to represent passengers and their getting on and off information in each bus stop. We aim to solve problems in bus operation situations by Visualization.


Time-Dependent Pedestrian Routing to Avoid Rain

In the cases of scattered rain, pedestrians try to find an optimal route to avoid walking under the rain and getting wet. Depending on the current available services, pedestrians can know where in the city is raining at the moment, however, it is difficult to know exactly which route would be optimal to avoid getting wet while walking.

In this research, we try to estimate the optimal route for the pedestrians in a rainy weather, by calculating a rainy cost in a time-constraint extension to shortest route algorithm.


Investigation of Road Surface Properties using Automatic Driving Vehicle Sensor Data

Automatic driving vehicles have many sensors to check the surrounding situation during autopilot. In this research, we aim at automatic detection of road surface deterioration based on automatic driving vehicle sensor data.

This research looks at the fact that many automated driving vehicles will run through the streets in the future.