Research Topics

"Heterogeneous no-lane traffic" is traffic in which various types of vehicles, such as motorcycles, passenger cars, trucks, and three-wheeled vehicles, travel along a road without lanes. We are working on the following topics for such traffic, e.g.,

• Search for the order pattern (group/herd) created by each vehicle type

• Assessing the risk of traffic congestion caused by such groups

• Creation of a model that reproduces the "groups" and "other randomized parts" of the traffic in the simulation.

Through these topics, we aim to achieve the following,

• Realize a simulation that can reproduce traffic characteristics more accurately

• Clarify the phenomenon of uneven distribution (segregation) of agents with different characteristics (vehicles/people/etc.).

Automated systems, such as automated driving, are useful because they reduce our effort and compensate for our mistakes. Advances in fields such as sensing, communication, computing, and machine learning have led these automated systems to surpass humans in the quality, quantity, and speed of their cognition and judgment in many areas.

We, humans, face the following problems in the face of such automated systems.

• Risk-taking behavior to counteract the increased safety afforded by the system

• Inappropriate use of the system due to overconfidence/unconfidence

• It takes time to pay attention to sudden contingencies.

• Cognitive biases can lead to irrational decisions

On the other hand, the system also has the following problems

• Cognition, judgment, and operation are still partially beyond the reach of humans.

• Insufficient/inappropriate communication between the system and the human.

To address these issues, we are conducting the following research, focusing on the relationship between human beings and driver assistance technologies/automated driving systems.

• Automatic driving that learns driving preferences and provides immediate feedback

• Information presentation methods enable the system to encourage desirable behavior that the driver does not recognize, such as preventing traffic congestion.