An Automated Machine Learning (AutoML) Method for Driving Distraction Detection Based on Lane-Keeping Performance

Abstract

With the enrichment of smartphones, driving distractions caused by phone usages have become a threat to driving safety. A promising way to mitigate driving distractions is to detect them and give real-time safety warnings. However, existing detection algorithms face two major challenges, low user acceptance caused by in-vehicle camera sensors, and uncertain accuracy of pre-trained models due to drivers' individual differences. Therefore, this study proposes a domain-specific automated machine learning (AutoML) to self-learn the optimal models to detect distraction based on lane-keeping performance data. The AutoML integrates the key modeling steps into an auto-optimizable pipeline, including knowledge-based feature extraction, feature selection by recursive feature elimination (RFE), algorithm selection, and hyperparameter auto-tuning by Bayesian optimization. An AutoML method based on XGBoost, termed AutoGBM, is built as the classifier for prediction and feature ranking. The model is tested based on driving simulator experiments of three driving distractions caused by phone usage: browsing short messages, browsing long messages, and answering a phone call. The proposed AutoGBM method is found to be reliable and promising to predict phone-related driving distractions, which achieves satisfactory results prediction, with a predictive power of 80% on group level and 90% on individual level accuracy. Moreover, the results also evoke the fact that each distraction types and drivers require different optimized hyperparameters values, which reconfirm the necessity of utilizing AutoML to detect driving distractions. The purposed AutoGBM not only produces better performance with fewer features; but also provides data-driven insights about system design.

All one needs to know about shared micromobility simulation: a complete survey

“…Secondly, factors about the vehicles itself. E-bike battery range [2][8] reflects the endurance mileage of the vehicle and it's crucial when allocating orders. When the e-bike is in used, the battery level will go down and when it's in a station, it can be recharged.…”

“…So, should be distinguished according to different points though it's the same factor. Soriguera F, et al [2] also care about whether the rental site is full, but they use it in a larger scale. They only care about full stations (marked 1) and empty stations (marked 0).…”

“…Although there are already a lot of researches about physical road [1], driving [2] [3] and aviation [4], the researches on aviation are still rare. Everything new is the unity of forward and tortuous.…”

“…Researches on the simulation of micromobility range from microscopic to macroscopic and increasingly attention is paid to human behavior and decision-making. Wu J L C [2] is one of the first researchers to simulate micromobility in 1996 using the Sun Microsystems Mobile IP model. Thanks to the invention of the Multi-Agent Transport Simulation Toolkit (MATSim), researches on the simulation of micromobility flourish.…”

“…Different from the papers mentioned above, the contribution of this paper is three-fold: (1) We propose an exhaustive taxonomy of concepts in the area of micromobility simulation, and present a comprehensive survey on micromobility simulation. (2) We provide a comprehensive review on micromobility simulation, how it is defined, and how it is related to or different from other similar simulation paradigms, such as transit simulation, ride and hail simulation, car simulation and pedestrian simulation. (3) We have compiled a list of challenges and future directions for research in micromobility simulation.…”