An official website of the United States government
ObjectiveThis study explores subjective and objective driving style similarity to identify how similarity can be used to develop driver-compatible vehicle automation. BackgroundSimilarity in the ways that interaction partners perform tasks can be measured subjectively, through questionnaires, or objectively by characterizing each agent’s actions. Although subjective measures have advantages in prediction, objective measures are more useful when operationalizing interventions based on these measures. Showing how objective and subjective similarity are related is therefore prudent for aligning future machine performance with human preferences. MethodsA driving simulator study was conducted with stop-and-go scenarios. Participants experienced conservative, moderate, and aggressive automated driving styles and rated the similarity between their own driving style and that of the automation. Objective similarity between the manual and automated driving speed profiles was calculated using three distance measures: dynamic time warping, Euclidean distance, and time alignment measure. Linear mixed effects models were used to examine how different components of the stopping profile and the three objective similarity measures predicted subjective similarity. ResultsObjective similarity using Euclidean distance best predicted subjective similarity. However, this was only observed for participants’ approach to the intersection and not their departure. ConclusionDeveloping driving styles that drivers perceive to be similar to their own is an important step toward driver-compatible automation. In determining what constitutes similarity, it is important to (a) use measures that reflect the driver’s perception of similarity, and (b) understand what elements of the driving style govern subjective similarity.
more »
« less
Real‐time driving style classification based on short‐term observations
Abstract Vehicle behaviour prediction provides important information for decision‐making in modern intelligent transportation systems. People with different driving styles have considerably different driving behaviours and hence exhibit different behaviour tendency. However, most existing prediction methods do not consider the different tendencies in driving styles and apply the same model to all vehicles. Furthermore, most of the existing driver classification methods rely on offline learning that requires a long observation of driving history and hence are not suitable for real‐time driving behaviour analysis. To facilitate personalised models that can potentially improve vehicle behaviour prediction, the authors propose an algorithm that classifies drivers into different driving styles. The algorithm only requires data from a short observation window and it is more applicable for real‐time online applications compared with existing methods that require a long term observation. Experiment results demonstrate that the proposed algorithm can achieve consistent classification results and provide intuitive interpretation and statistical characteristics of different driving styles, which can be further used for vehicle behaviour prediction.
more »
« less
- Award ID(s):
- 1932139
- PAR ID:
- 10570744
- Publisher / Repository:
- DOI PREFIX: 10.1049
- Date Published:
- Journal Name:
- IET Communications
- Volume:
- 16
- Issue:
- 12
- ISSN:
- 1751-8628
- Format(s):
- Medium: X Size: p. 1393-1402
- Size(s):
- p. 1393-1402
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Autonomous vehicles are cyber-physical systems that combine embedded computing and deep learning with physical systems to perceive the world, predict future states, and safely control the vehicle through changing environments. The ability of an autonomous vehicle to accurately predict the motion of other road users across a wide range of diverse scenarios is critical for both motion planning and safety. However, existing motion prediction methods do not explicitly model contextual information about the environment, which can cause significant variations in performance across diverse driving scenarios. To address this limitation, we proposeCASTNet: a dynamic, context-aware approach for motion prediction that (i) identifies the current driving context using a spatio-temporal model, (ii) adapts an ensemble of motion prediction models to fit the current context, and (iii) applies novel trajectory fusion methods to combine predictions output by the ensemble. This approach enables CASTNet to improve robustness by minimizing motion prediction error across diverse driving scenarios. CASTNet is highly modular and can be used with various existing image processing backbones and motion predictors. We demonstrate how CASTNet can improve both CNN-based and graph-learning-based motion prediction approaches and conduct ablation studies on the performance, latency, and model size for various ensemble architecture choices. In addition, we propose and evaluate several attention-based spatio-temporal models for context identification and ensemble selection. We also propose a modular trajectory fusion algorithm that effectively filters, clusters, and fuses the predicted trajectories output by the ensemble. On the nuScenes dataset, our approach demonstrates more robust and consistent performance across diverse, real-world driving contexts than state-of-the-art techniques.more » « less
-
Adaptive Cruise Control (ACC) has become increasingly popular in modern vehicles, providing enhanced driving safety, comfort, and fuel efficiency. However, predefined ACC settings may not always align with a driver's preferences, leading to discomfort and possible safety hazards. To address this issue, Personalized ACC (P-ACC) has been studied by scholars. However, existing research mostly relies on historical driving data to imitate driver styles, which ignores real-time feedback from the driver. To overcome this limitation, we propose a cloud-vehicle collaborative P-ACC framework, which integrates real-time driver feedback adaptation. This framework consists of offline and online modules. The offline module records the driver's naturalistic car-following trajectory and uses inverse reinforcement learning (IRL) to train the model on the cloud. The online module utilizes the driver's real-time feedback to update the driving gap preference in real-time using Gaussian process regression (GPR). By retraining the model on the cloud with the driver's takeover trajectories, our approach achieves incremental learning to better match the driver's preference. In human-in-the-loop (HuiL) simulation experiments, the proposed framework results in a significant reduction of driver intervention in automatic control systems, up to 70.9%.more » « less
-
ABSTRACT Traffic accidents have become a major concern for governments, organizations and individuals worldwide due to the material and moral losses they cause. It is possible to reduce this concern by taking into account the research conducted by relevant institutions and organizations in this field. The main objective of this study is to categorize traffic accidents according to driver violation types and analyse them using machine learning algorithms and feature sensitivity to identify the most influential variables in each category. For this purpose, traffic accident reports that occurred in Erzurum province in the last 1 year were used to categorize and classify driver violation behaviour types. Five different machine learning algorithms, namely k‐nearest neighbour, support vector machines, naive Bayes, multilayer perception and random forest, were used to examine the success performance of the classification. Among these, 91% successful classification was obtained with the random forest algorithm. Based on the classification obtained from this algorithm, sensitivity analysis was used to reveal the variables that most affect each violation category. The results of the analysis revealed that driver age and vehicle type were the most influential variables for many types of violations. Thanks to this study, the problems were clearly identified by going into the details of driver violation behaviours. At the end of the study, measures to reduce driver violation behaviours were proposed. If the recommendations that can reduce driver behaviour are taken into consideration by transportation authorities and policy makers, traffic accidents can be significantly reduced.more » « less
-
Abstract For simulation to be an effective tool for the development and testing of autonomous vehicles, the simulator must be able to produce realistic safety-critical scenarios with distribution-level accuracy. However, due to the high dimensionality of real-world driving environments and the rarity of long-tail safety-critical events, how to achieve statistical realism in simulation is a long-standing problem. In this paper, we develop NeuralNDE, a deep learning-based framework to learn multi-agent interaction behavior from vehicle trajectory data, and propose a conflict critic model and a safety mapping network to refine the generation process of safety-critical events, following real-world occurring frequencies and patterns. The results show that NeuralNDE can achieve both accurate safety-critical driving statistics (e.g., crash rate/type/severity and near-miss statistics, etc.) and normal driving statistics (e.g., vehicle speed/distance/yielding behavior distributions, etc.), as demonstrated in the simulation of urban driving environments. To the best of our knowledge, this is the first time that a simulation model can reproduce the real-world driving environment with statistical realism, particularly for safety-critical situations.more » « less
