In this study, assuming an in-vehicle environment during automatic driving, we investigated the effectiveness of a sound environment where nature sounds are being played, which are considered to have no significant differences in passenger preferences, with the aim of improving comfort, including not only maintaining arousal but also aspects such as mental/physical activation. In the first experiment, two types of driving movies were prepared in consideration of the influence of visual information, and a subjective evaluation experiment of nature sounds was conducted to determine what kind of nature sound can improve comfort regardless of the driving environment. As a result, it was found that the sound source that can contribute to improving comfort and arousal in the in-vehicle environment is birdsong, and that a sound source with a particularly high frequency component is more appropriate. In the second experiment, the contribution of playing the nature sound, which was considered optimal in the first experiment, to maintaining driver arousal was determined. The results suggest that the five-second nature sound can speed up reaction time and reduce reaction delays, and that the effect may be particularly significant at 10-second interval presentations or at 30-second interval presentations with relatively high levels.

For the practical implementation of autonomous driving, it is required that a single operator remotely monitors multiple vehicles and performs recovery operations as needed. To achieve this, it is essential to reduce the burden on operators during recovery operations. This study evaluates the impact of information presentation of remote monitoring systems on the operator’s workload and work efficiency. The remote monitoring systems used in previous autonomous driving demonstration experiments can be classified into two types based on the differences in their expression of the recognition and judgment of autonomous driving systems. We conducted experiments simulating these two systems and analyzed the characteristics and challenges of each system. The results indicated that the system providing system-level information allowed operators to pay attention to objects that humans typically do not focus on, while it might increase annoyance in high-traffic situations. On the other hand, in the system that presents information concisely with diagrams, although there are objects that are difficult to notice, if the provided information is appropriate for situation understanding, it can guide operators’ attention and potentially reduce the operation time.

This study utilizes science fiction prototyping and Live Action Role Playing (LARP) to enhance the strategic planning processes within corporations and governments. Participants are encouraged to actively imagine and shape future societal outcomes by deeply engaging in scifi-based LARP, which emphasizes psychological immersion within physical spaces. The evaluation by workshop participants through interviews suggested that this method allowed participants to embody characters and explore complex relationships, fostering a comprehensive understanding of future scenarios. The immersive experience not only promotes alignment among diverse stakeholders but also enables detailed discussions on the potential societal impacts of proposed solutions. Post-project observations indicate that the benefits of this immersive approach are likely to have lasting effects.

In this study, we propose an overall self-instructional training support system and tested the effectiveness of a self-instructional text presentation interface using a speech bubble. While the automated generation of self-instructional texts could make the training more accessible, previous research shows that if the texts are not self-created, the effectiveness of the training decreases. To overcome this, we design an interface that displays a real-time video of the user’s face with superimposed speech bubbles containing self-instructional text. Through online experiments, we demonstrated that the interface enhances users’ internalization of the generated text, potentially improving the effectiveness of computer-assisted self-instructional training.