Real time detection of driver attention: Emerging solutions based on robust iconic classifiers and dictionary of poses

Real time monitoring of driver attention by computer vision techniques is a key issue in the development of advanced driver assistance systems. While past work mostly focused on structured feature-based approaches, characterized by high computational requirements, emerging technologies based on iconic classifiers recently proved to be good candidates for the implementation of accurate and real-time solutions, characterized by simplicity and automatic fast training stages.In this work the combined use of binary classifiers and iconic data reduction, based on Sanger neural networks, is proposed, detailing critical aspects related to the application of this approach to the specific problem of driving assistance. In particular it is investigated the possibility of a simplified learning stage, based on a small dictionary of poses, that makes the system almost independent from the actual user.On-board experiments demonstrate the effectiveness of the approach, even in case of noise and adverse light conditions. Moreover the system proved unexpected robustness to various categories of users, including people with beard and eyeglasses. Temporal integration of classification results, together with a partial distinction among visual distraction and fatigue effects, make the proposed technology an excellent candidate for the exploration of adaptive and user-centered applications in the automotive field.     

Incorporating the extended theory of planned behavior in a school travel mode choice model: a case study of Shaoxing,China

Travel to and from school can have social, economic, and environmental implications for students and their parents. Therefore, understanding school travel mode choice behavior is essential to find policy-oriented approaches to optimizing school travel mode share. Recent research suggests that psychological factors of parents play a significant role in school travel mode choice behavior and the Multiple Indicators and Multiple Causes (MIMIC) model has been used to test the effect of psychological constructs on mode choice behavior. However, little research has used a systematic framework of behavioral theory to organize these psychological factors and investigate their internal relationships. This paper proposes an extended theory of planned behavior (ETPB) to delve into the psychological factors caused by the effects of adults’ cognition and behavioral habits and explores the factors’ relationship paradigm. A theoretical framework of travel mode choice behavior for students in China is constructed. We established the MIMIC model that accommodates latent variables from ETPB. We found that not all the psychological latent variables have significant effects on school travel mode choice behavior, but habit can play an essential role. The results provide theoretical support for demand policies for school travel.     

A Pareto-improving hybrid rationing and pricing policy with multiclass network equilibria

This paper extends the work on Pareto-improving hybrid rationing and pricing policy for general road networks by considering heterogeneous users with different values of time. Mathematical programming models are proposed to find a multiclass Pareto-improving pure road space rationing scheme (MPI-PR) and multiclass hybrid rationing and pricing schemes (MHPI and MHPI-S). A numerical example with a multimodal network is provided for comparing both the efficiency and equity of the three proposed policies. We discover that MHPI-S can achieve the largest reduction in total system delay, MHPI can induce the least spatial inequity and MHPI-S is a progressive policy which is appealing to policy makers. Furthermore, numerical results reveal that different classes of users react differently to the same hybrid policies and multiclass Pareto-improving hybrid schemes yield less delay reduction when compared to their single-class counterparts.     

You (expect to) get what you pay for: A system approach to delay, fare, and complaints

In this paper, an analytical framework integrating delay, fare, and complaints with passenger air travel has been laid out. Examining aggregate monthly data for US domestic air travel, we have identified causal relationships among fare, complaints, and levels of delay. An analytical framework is proposed that formalizes these relationships in an integrated manner. This integrated framework is then estimated in a set of simultaneous equations by using 118 months of data from January 1997 to October 2006. Results show that complaints are influenced by levels of delays. However, complaints are positively influenced by average yield. These findings lead us to support the central hypothesis that complaints are responsive to levels of delays, but they tend to vary according to fare. That is, air travelers are less likely to complain in return for lower fares, even when faced with the same or even higher levels of delays. These findings have important policy implications, including the passengers’ bill of rights and regulator’s choice between market and operational performances.     

埠阳庄大桥复杂地质条件下的桩基施工

埠阳庄大桥实际地质状况复杂,重点介绍该桥桩基施工中的相关准备工作以及对问题的分析和处理措施,为同类地质条件下梁桩基施工提供借鉴经验。     

注汽管道承压能力评估和实物试验验证

基于失效评估图技术对某注汽管道进行了承压能力评价,给出了管线在不同腐蚀壁厚情况下的极限承压能力和极限裂纹尺寸。并利用实物爆破试验验证了评价方法的安全性和有效性,通过对爆破断口进行宏观和微观分析,发现管段的开裂属于脆性开裂。     

长距离密闭输油管道的水击数值模拟

在长距离密闭输油管道中间泵站断电停运的情况下,采用特征线法对全线压力变化情况进行了数值模拟。结果表明:越是靠近下游的泵站停运后,越能更快地形成越站输送,出现进站压力小于泵吸入压力的时间越晚,管道的最大压力越大,管道出现液柱分离的机会越小等。     

Locating emergency vehicles with an approximate queuing model and a meta-heuristic solution approach

In this paper, the location of emergency service (ES) vehicles is studied on fully connected networks. Queuing theory is utilized to obtain the performance metrics of the system. An approximate queuing model the (AQM) is proposed. For the AQM, different service rate formulations are constructed. These formulations are tested with a simulation study for different approximation levels. A mathematical model is proposed to minimize the mean response time of ES systems based on AQM. In the model, multiple vehicles are allowed at a single location. The objective function of the model has no closed form expression. A genetic algorithm is constructed to solve the model. With the help of the genetic algorithm, the effect of assigning multiple vehicles on the mean response time is reported.     

Human-like autonomous car-following model with deep reinforcement learning

This study proposes a framework for human-like autonomous car-following planning based on deep reinforcement learning (deep RL). Historical driving data are fed into a simulation environment where an RL agent learns from trial and error interactions based on a reward function that signals how much the agent deviates from the empirical data. Through these interactions, an optimal policy, or car-following model that maps in a human-like way from speed, relative speed between a lead and following vehicle, and inter-vehicle spacing to acceleration of a following vehicle is finally obtained. The model can be continuously updated when more data are fed in. Two thousand car-following periods extracted from the 2015 Shanghai Naturalistic Driving Study were used to train the model and compare its performance with that of traditional and recent data-driven car-following models. As shown by this study’s results, a deep deterministic policy gradient car-following model that uses disparity between simulated and observed speed as the reward function and considers a reaction delay of 1 s, denoted as DDPGvRT, can reproduce human-like car-following behavior with higher accuracy than traditional and recent data-driven car-following models. Specifically, the DDPGvRT model has a spacing validation error of 18% and speed validation error of 5%, which are less than those of other models, including the intelligent driver model, models based on locally weighted regression, and conventional neural network-based models. Moreover, the DDPGvRT demonstrates good capability of generalization to various driving situations and can adapt to different drivers by continuously learning. This study demonstrates that reinforcement learning methodology can offer insight into driver behavior and can contribute to the development of human-like autonomous driving algorithms and traffic-flow models.     

Compliance with MARPOL Annex VI regulation 14 by ships in the Gulf of Guinea sub-region: Issues,challenges and opportunities

The negative environmental and health impacts association with high sulphur dioxide emissions from shipboard machineries have been raised by various stakeholders within the marine transportation sector. It is against this backdrop that the International Maritime Organisation under the MARPOL Annex VI regulation 14 has capped sulphur emission to 0.1% for Sulphur Emission Control Areas and 0.5% for the other shipping nations. However, ship owners in the Gulf of Guinea (GoG) sub-region are facing multitudes of challenges in meeting up with this new IMO regulation. This paper aims to identify the main barriers hampering effective compliance to this new regulation by ships operating in the GoG, rank the barriers, and then discuss the possible opportunities that may arise as a result of addressing the challenges. To identify the main barriers, experts with several years of experience in the maritime industry from Ghana and Cameroun were used while multi-criteria decision-making (MCDM) method combining analytic hierarchy process (AHP) and technique for order preference by similarity to ideal solution (TOPSIS) was employed to rank the barriers. Other methods such as fuzzy AHP (FAHP), rank-order centroid (ROC) and TOPSIS were combined to validate the result of the study. The findings indicate that lack of infrastructure, lack of comprehensive marine air pollution laws and high capital and operational costs of sulphur reduction solutions emerged as the top three ranked barriers. The findings of this study can be useful to ship owners and policy makers in dealing with the issues of marine air pollution.