Identification and Classification of Human Performance related Challenges during Remote Driving
Remote driving of vehicles is gaining in importance in the transportation sector, especially when Automated Driving Systems (ADSs) reach the limits of their system boundaries. This study investigates the challenges faced by human Remote Drivers (RDs) during remote driving, particularly focusing on the identification and classification of human performance-related challenges through a comprehensive analysis of real-world remote driving data Las Vegas. For this purpose, a total of 183 RD performance-related Safety Driver (SD) interventions were analyzed and classified using an introduced severity classification. As it is essential to prevent the need for SD interventions, this study identified and analyzed harsh driving events to detect an increased likelihood of interventions by the SD. In addition, the results of the subjective RD questionnaire are used to evaluate whether the objective metrics from SD interventions and harsh driving events can also be confirmed by the RDs and whether additional challenges can be uncovered. The analysis reveals learning curves, showing a significant decrease in SD interventions as RD experience increases. Early phases of remote driving experience, especially below 200 km of experience, showed the highest frequency of safety-related events, including braking late for traffic signs and responding impatiently to other traffic participants. Over time, RDs follow defined rules for improving their control, with experience leading to less harsh braking, acceleration, and steering maneuvers. The study contributes to understanding the requirements of RDS, emphasizing the importance of targeted training to address human performance limitations. It further highlights the need for system improvements to address challenges like latency and the limited haptic feedback replaced by visual feedback, which affect the RDs' perception and vehicle control.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Human Body Pose Estimation for Gait Identification: A Comprehensive Survey of Datasets and Models
Person identification is a problem that has received substantial attention, particularly in security domains. Gait recognition is one of the most convenient approaches enabling person identification at a distance without…
Gait IdentificationGait RecognitionPerson IdentificationPose Estimation+1Overview of PlantCLEF 2023: Image-based Plant Identification at Global Scale
The world is estimated to be home to over 300,000 species of vascular plants. In the face of the ongoing biodiversity crisis, expanding our understanding of these species is crucial for the advancement of human civilizat…
M2H-GAN: A GAN-based Mapping from Machine to Human Transcripts for Speech Understanding
Deep learning is at the core of recent spoken language understanding (SLU) related tasks. More precisely, deep neural networks (DNNs) drastically increased the performances of SLU systems, and numerous architectures have…
Generative Adversarial NetworkSpoken Language UnderstandingFD-GAN: Pose-guided Feature Distilling GAN for Robust Person Re-identification
Person re-identification (reID) is an important task that requires to retrieve a person's images from an image dataset, given one image of the person of interest. For learning robust person features, the pose variation o…
Generative Adversarial NetworkPerson Re-IdentificationComplex Word Identification: Challenges in Data Annotation and System Performance
This paper revisits the problem of complex word identification (CWI) following up the SemEval CWI shared task. We use ensemble classifiers to investigate how well computational methods can discriminate between complex an…
Complex Word IdentificationGeneral Classification