No abstract
Today, and possibly for a long time to come, the full driving task is too complex an activity to be fully formalized as a sensing-acting robotics system that can be explicitly solved through model-based and learning-based approaches in order to achieve full unconstrained vehicle autonomy. Localization, mapping, scene perception, vehicle control, trajectory optimization, and higher-level planning decisions associated with autonomous vehicle development remain full of open challenges. This is especially true for unconstrained, real-world operation where the margin of allowable error is extremely small and the number of edge-cases is extremely large. Until these problems are solved, human beings will remain an integral part of the driving task, monitoring the AI system as it performs anywhere from just over 0% to just under 100% of the driving. The governing objectives of the MIT Advanced Vehicle Technology (MIT-AVT) study are to (1) undertake large-scale real-world driving data collection that includes high-definition video to fuel the development of deep learning based internal and external perception systems, (2) gain a holistic understanding of how human beings interact with vehicle automation technology by integrating video data with vehicle state data, driver characteristics, mental models, and self-reported experiences with technology, and (3) identify how technology and other factors related to automation adoption and use can be improved in ways that save lives. In pursuing these objectives, we have instrumented 23 Tesla Model S and Model X vehicles, 2 Volvo S90 vehicles, 2 Range Rover Evoque, and 2 Cadillac CT6 vehicles for both long-term (over a year per driver) and medium term (one month per driver) naturalistic driving data collection. Furthermore, we are continually developing new methods for analysis of the massive-scale dataset collected from the instrumented vehicle fleet. The recorded data streams include IMU, GPS, CAN messages, and high-definition video streams of the driver face, the driver cabin, the forward roadway, and the instrument cluster (on select vehicles). The study is on-going and growing. To date, we have 122 participants, 15,610 days of participation, 511,638 miles, and 7.1 billion video frames. This paper presents the design of the study, the data collection hardware, the processing of the data, and the computer vision algorithms currently being used to extract actionable knowledge from the data. 01 231 4523 67 89 8 %& 'ÿ )*+ ,-,.,*/ÿ 0123 45 1 '142-,5 ,67ÿ 8+ *97 :; <=>ÿ @AB; CDÿ ; AE =F; GH ÿIJ KFL; M NM OFB; ÿ =F>DH ÿPQR SPT ULM VGLDH ÿPWW XGCM NY GDH ÿWZ [M Y GDÿ =LM VGBH ÿQPPR SI\ XM =GAÿ ] LF@GDH ÿJPPÿa b b a cd :; <=>ÿ =F; Fÿ NAY Y GN; M ABÿ M Dÿ ABeAM Bef ÿ :; F; M D; M NDÿ
This work may have important implications for attention management in the context of the increasing prevalence of in-vehicle demands as well as of vehicle automation.
This study examined whether the effect of subsidiary tasks on driving performance can be predicted from stationary (static) testing. Alternative methods for assessing the performance of drivers during their use of in-vehicle information systems were examined. These methods included static testing in stationary vehicles, as well as dynamic, on-road testing. The measures that were obtained from static tests were evaluated in terms of how well they could predict measures obtained from driving performance during on-road testing (which included concurrent use of secondary information systems). The results indicated that measures obtained in static test settings were highly correlated with corresponding measures obtained from on-road performance testing. Also, the regression slopes and intercepts of on-road driving performance measures when predicting on-road lane deviations were more closely approximated under static conditions which required drivers to attend to concurrent, driving-like activities as well as secondary tasks. It is suggested that tests using multiple measures, together with a multivariate model, may hold the most promise for predicting eventual on-road driver performance in the early development of in-vehicle information systems. Although additional research and theoretical development are required, this research demonstrates that a set of relatively simple measures obtained analytically and from static testing can be used even now to help guide early system development.
Summary: This analysis identified the underlying dimensions of driver performance using data obtained from drivers engaged in secondary manual tasks. Randomly chosen subjects balanced for age and gender used one of five advanced navigation and communication systems while driving on a closed roadway. Fifteen driver performance variables were averaged and standardized across subjects for 79 tasks. There were high correlations between all variables. Principal Component Analysis (PCA) found that the vector of loadings defining the first principal component (PC1) was positive for all 15 variables, accounting for 61 percent of the total variation across all tasks. It is interpreted as overall driver demand. PC2 loaded with one sign on event detection and response variables, but opposing sign on visual-manual workload variables. It identified tasks making drivers more inattentive to outside events than expected, given a task's visual-manual workload, and accounted for 17 percent of total variation. It is interpreted as low-workloadbut-high-inattentiveness. PC3 had loadings of opposing sign for peripheral vs. central event variables (5 percent of total variation). It is interpreted as peripheral insensitivity. The first three components together accounted for 83 percent of total variation, which is deemed substantial. Thus most of the information available through the 15 original variables can be summarized by only three PC variables. Because the vectors of loadings defining the components are orthogonal to each other as defined by PCA, no single variable by itself can capture all the important variations in driver performance during secondary manual tasks. A multivariate design and analysis is required.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.