Modeling takeover time based on non-driving-related task attributes in highly automated driving.

This study aims to investigate the effects of non-driving-related tasks (NDRTs) on the transition of control in highly automated driving (HAD) by investigating the effects of NDRT physical, visual, and cognitive attributes during transition of control. A conceptual model of the takeover process is proposed by dividing this process into motor and mental reactions. A laboratory experiment was conducted to evaluate the effects of each NDRT attribute on the corresponding stage of the process of taking over control. A prediction model was developed using the results of multiple linear regression analysis. Additionally, a validation experiment with nine NDRTs and a baseline condition was conducted to determine the extent to which the developed model explains the takeover time for each NDRT condition. The results showed that the timing aspects of the transition of control in HAD largely consist of participant motor reactions that are affected by the physical attributes of NDRTs.

Early-Stage Lung Cancer Diagnosis by Deep Learning-Based Spectroscopic Analysis of Circulating Exosomes.

Lung cancer has a high mortality rate, but an early diagnosis can contribute to a favorable prognosis. A liquid biopsy that captures and detects tumor-related biomarkers in body fluids has great potential for early-stage diagnosis. Exosomes, nanosized extracellular vesicles found in blood, have been proposed as promising biomarkers for liquid biopsy. Here, we demonstrate an accurate diagnosis of early-stage lung cancer, using deep learning-based surface-enhanced Raman spectroscopy (SERS) of the exosomes. Our approach was to explore the features of cell exosomes through deep learning and figure out the similarity in human plasma exosomes, without learning insufficient human data. The deep learning model was trained with SERS signals of exosomes derived from normal and lung cancer cell lines and could classify them with an accuracy of 95%. In 43 patients, including stage I and II cancer patients, the deep learning model predicted that plasma exosomes of 90.7% patients had higher similarity to lung cancer cell exosomes than the average of the healthy controls. Such similarity was proportional to the progression of cancer. Notably, the model predicted lung cancer with an area under the curve (AUC) of 0.912 for the whole cohort and stage I patients with an AUC of 0.910. These results suggest the great potential of the combination of exosome analysis and deep learning as a method for early-stage liquid biopsy of lung cancer.

Effects of visual complexity of in-vehicle information display: Age-related differences in visual search task in the driving context.

We aimed to investigate the effects of the visual complexity of in-vehicle information display and driver's age in a driving context. A driving simulator study was conducted where participants performed visual search tasks at different visual complexity levels while driving. Two groups were included, 20 younger drivers (mean age = 28.75 years) and 14 older drivers (mean age = 54.87 years). Older drivers were found to be more vulnerable to the effects of increased visual complexity when performing a visual search task. The task completion time of the younger group increased by about 20% (from 7.69 s to 9.30 s), while the older group increased by about 47% (from 8.92 s to 13.14 s). Further, the driving performance of the older group deteriorated, unlike the younger group. The subjective workload score supported the results of the objective performance measures. These differences can be explained by glance behavior. The total off-road glance duration of older drivers was longer than that of younger drivers, but the average off-road glance duration of younger drivers was longer. In other words, older drivers have a more conservative strategy when dealing with increased visual complexity in a driving context so as not to affect their driving. The findings of this study show that the visual complexity level has a significant effect on driving behaviors, especially in older drivers, which provides insights for designing in-vehicle information displays.

The effects of takeover request modalities on highly automated car control transitions.

This study investigated the influences of takeover request (TOR) modalities on a drivers' takeover performance after they engaged in non-driving related (NDR) tasks in highly automated driving (HAD). Visual, vibrotactile, and auditory modalities were varied in the design of the experiment under four conditions: no-task, phone conversation, smartphone interaction, and video watching tasks. Driving simulator experiments were conducted to analyze the drivers' take-over performance by collecting data during the transition time of re-engaging control of the vehicle, the time taken to be on the loop, and time taken to be physically ready to drive. Data were gathered on the perceived usefulness, safety, satisfaction, and effectiveness for each TOR based on a self-reported questionnaire. Takeover and hands-on times varied considerably, as shown by high standard deviation values between modalities, especially for phone conversations and smartphone interaction tasks. Moreover, it was found that participants failed to take over control of the vehicle when they were given visual TORs for phone conversation and smartphone interaction tasks. The perceived safety and satisfaction varied for the NDR task. Results from the statistical analysis showed that the NDR task significantly influenced the takeover time, but there was no significant interaction effect between the TOR modalities and the NDR task. The results could potentially be applied to the design of safe and efficient transitions of highly controlled, automated driving, where drivers are enabled to engage in NDR tasks.

Smartphone form factors: Effects of width and bottom bezel on touch performance, workload, and physical demand.

This study aimed at investigating the effect of two smartphone form factors (width and bottom bezel) on touch behaviors with one-handed interaction. User experiments on tapping tasks were conducted for four widths (67, 70, 72, and 74 mm) and five bottom bezel levels (2.5, 5, 7.5, 10, and 12.5 mm). Task performance, electromyography, and subjective workload data were collected to examine the touch behavior. The success rate and task completion time were collected as task performance measures. The NASA-TLX method was used to observe the subjective workload. The electromyogram signals of two thumb muscles, namely the first dorsal interosseous and abductor pollicis brevis, were observed. The task performances deteriorated with increasing width level. The subjective workload and electromyography data showed similar patterns with the task performances. The task performances of the bottom bezel devices were analyzed by using three different evaluation criteria. The results from these criteria indicated that tasks became increasingly difficult as the bottom bezel level decreased. The results of this study provide insights into the optimal range of smartphone form factors for one-handed interaction, which could contribute to the design of new smartphones.

How we can measure the non-driving-task engagement in automated driving: Comparing flow experience and workload.

In automated driving, a driver can completely concentrate on non-driving-related tasks (NDRTs). This study investigated the flow experience of a driver who concentrated on NDRTs and tasks that induce mental workload under conditional automation. Participants performed NDRTs under different demand levels: a balanced demand-skill level (fit condition) to induce flow, low-demand level to induce boredom, and high-demand level to induce anxiety. In addition, they performed the additional N-Back task, which artificially induces mental workload. The results showed participants had the longest reaction time when they indicated the highest flow score, and had the longest gaze-on time, road-fixation time, hands-on time, and take-over time under the fit condition. Significant differences were not observed in the driver reaction times in the fit condition and the additional N-Back task, indicating that performing NDRTs that induce a high flow experience could influence driver reaction time similar to performing tasks with a high mental workload.

Vibrotactile perception assessment for a haptic interface on an antigravity suit.

Haptic technology is used in various fields to transmit information to the user with or without visual and auditory cues. This study aimed to provide preliminary data for use in developing a haptic interface for an antigravity (anti-G) suit. With the structural characteristics of the anti-G suit in mind, we determined five areas on the body (lower back, outer thighs, inner thighs, outer calves, and inner calves) on which to install ten bar-type eccentric rotating mass (ERM) motors as vibration actuators. To determine the design factors of the haptic anti-G suit, we conducted three experiments to find the absolute threshold, moderate intensity, and subjective assessments of vibrotactile stimuli. Twenty-six fighter pilots participated in the experiments, which were conducted in a fixed-based flight simulator. From the results of our study, we recommend 1) absolute thresholds of ∼11.98-15.84 Hz and 102.01-104.06 dB, 2) moderate intensities of 74.36 Hz and 126.98 dB for the lower back and 58.65 Hz and 122.37 dB for either side of the thighs and calves, and 3) subjective assessments of vibrotactile stimuli (displeasure, easy to perceive, and level of comfort). The results of this study will be useful for the design of a haptic anti-G suit.

Assessment model for perceived visual complexity of automotive instrument cluster.

This research proposes an assessment model for quantifying the perceived visual complexity (PVC) of an in-vehicle instrument cluster. An initial study was conducted to investigate the possibility of evaluating the PVC of an in-vehicle instrument cluster by estimating and analyzing the complexity of its individual components. However, this approach was only partially successful, because it did not take into account the combination of the different components with random levels of complexity to form one visual display. Therefore, a second study was conducted focusing on the effect of combining the different components. The results from the overall research enabled us to suggest a basis for quantifying the PVC of an in-vehicle instrument cluster based both on the PVCs of its components and on the integration effect.

Interface methods for using intranet portal organizational memory information system.

In this paper, an intranet portal is considered as an information infrastructure (organizational memory information system, OMIS) supporting organizational learning. The properties and the hierarchical structure of information and knowledge in an intranet portal OMIS was identified as a problem for navigation tools of an intranet portal interface. The problem relates to navigation and retrieval functions of intranet portal OMIS and is expected to adversely affect user performance, satisfaction, and usefulness. To solve the problem, a conceptual model for navigation tools of an intranet portal interface was proposed and an experiment using a crossover design was conducted with 10 participants. In the experiment, a separate access method (tabbed tree tool) was compared to an unified access method (single tree tool). The results indicate that each information/knowledge repository for which a user has a different structural knowledge should be handled separately with a separate access to increase user satisfaction and the usefulness of the OMIS and to improve user performance in navigation.