An inference engine for smartphones to preprocess data and detect stationary and transportation modes

A smartphone can be utilized as a cost-effective device for the purposes of intelligent transportation system. To detect the movement and the stationary statuses in the motorized and non-motorized modes, this study develops a new inference engine, including two sets of rules. The first sets of rules are defined by the related thresholds on the features of smartphone sensors while the second sets are extracted from the human knowledge to improve the results of the first rules. The experimental results reveal that by utilizing Inertial Measurement Unit (IMU) sensors in the proposed inference engine, it is possible to save 40% energy in comparison with the previous research. Moreover, this engine increases the accuracy of the motorized mode detection to 95.2% and determines the stationary states in motorized mode with 97.1% accuracy.     

Field tests of a dynamic green driving strategy based on inter-vehicle communication

This paper presents the design and results for field tests regarding the environmental benefits in stop-and-go traffic of an algorithmic green driving strategy based on inter-vehicle communication (IVC), which was proposed in Yang and Jin (2014). The green driving strategy dynamically calculates advisory speed limits for vehicles equipped with IVC devices so as to smooth their speed profiles and reduce their emissions and fuel consumption. For the field tests, we develop a smartphone-based IVC system, in which vehicles’ speeds and locations are collected by GPS and accelerometer sensors embedded in smartphones, and communications among vehicles are enabled by specially designed smartphone applications, a central server, and 4G cellular networks. Six field tests are carried out on an uninterrupted ring road under slow or fast stop-and-go traffic conditions. We compare the performances of three alternatives: no green driving, heuristic green driving, and the IVC-based algorithmic green driving. Results show that heuristic green driving has better smoothing and environmental effects than no green driving, but the IVC-based algorithmic green driving outperforms both. In the future, we are interested in field tests under more realistic traffic conditions.     

Estimation of driving style in naturalistic highway traffic using maneuver transition probabilities

Accurately estimating driving styles is crucial to designing useful driver assistance systems and vehicle control systems for autonomous driving that match how people drive. This paper presents a novel way to identify driving style not in terms of the durations or frequencies of individual maneuver states, but rather the transition patterns between them to see how they are interrelated. Driving behavior in highway traffic was categorized into 12 maneuver states, based on which 144 (12 × 12) maneuver transition probabilities were obtained. A conditional likelihood maximization method was employed to extract typical maneuver transition patterns that could represent driving style strategies, from the 144 probabilities. Random forest algorithm was adopted to classify driving styles using the selected features. Results showed that transitions concerning five maneuver states – free driving, approaching, near following, constrained left and right lane changes – could be used to classify driving style reliably. Comparisons with traditional methods were presented and discussed in detail to show that transition probabilities between maneuvers were better at predicting driving style than traditional maneuver frequencies in behavioral analysis.     

An android based ecodriving assistance system to improve safety and efficiency of internal combustion engine passenger cars

Recent improvements of communication technologies leads to several innovations in road vehicles energy consumption. As an example, several ecodriving applications already appeared on all smartphone application markets. Using embedded smartphone signals, such applications provide real time feedback to drivers according to their performances. However most of these applications does not take into account upcoming events such as curves, slopes or crossings to advise the driver on the best actions to undertake to lower energy consumption. Furthermore, they do not analyze data coming from vehicle sensors. In this paper, we present an android application, developed within the FP7 European project ecoDriver, which provides several innovative properties: advice according to upcoming events, a real time evaluation of the driving behavior, the analysis of past actions, an interface with OBD2 connector and some more. This paper further develops the complete architecture and links between each innovative function. Future works will concentrate on integrating image processing in this application in order to detect the possible presence of a front vehicle.     

Study on the framework of hybrid collision warning system using loop detectors and vehicle information

Safety warning systems generally operate based on information from sensors attached to individual vehicles. Various types of data used for collision risk calculation can be categorized into two types, microscopic or macroscopic, depending on how the sensors collect the information of traffic state. Most collision warning systems use only either of these types of data, but they all have limitations imposed by the data, such as requirement of high installation cost and high market penetration rate of devices. In order to overcome these limits, we propose a collision warning system that utilizes the integrated information of macroscopic data and microscopic data, from loop detectors and smartphones respectively. The proposed system is evaluated by simulating a real vehicle trip based on the NGSIM data. We compare the results against collision warning systems based on macroscopic data from infrastructure and microscopic data from Vehicle-to-Vehicle information. The analysis of three systems shows two findings that (a) ICWS (Infrastructure-based Collision Warning System) is inadequate for immediate collision warning system and (b) VCWS (V2V communication based Collision Warning System) and HCWS (Hybrid Collision Warning System) produce collision warning at very similar timing, even with different behavior of individual drivers. Advantages of HCWS are that it can be directly applied to existing system with small additional cost, because data of loop detector are already available to be used in Korea and smartphones are widely spread. Also, the computation power distributed to each individual smartphone greatly increases the efficiency of the system by distributing the computation resources and load.     

An eigenmodel for iterative line planning,timetabling and vehicle scheduling in public transportation

Planning a public transportation system is a multi-objective problem which includes among others line planning, timetabling, and vehicle scheduling. For each of these planning stages, models are known and advanced solution techniques exist. Some of the models focus on costs, others on passengers’ convenience. Setting up a transportation system is usually done by optimizing each of these stages sequentially.In this paper we argue that instead of optimizing each single step further and further it would be more beneficial to consider the whole process in an integrated way. To this end, we develop and discuss a generic, bi-objective model for integrating line planning, timetabling, and vehicle scheduling. We furthermore propose an eigenmodel which we apply for these three planning stages and show how it can be used for the design of iterative algorithms as heuristics for the integrated problem. The convergence of the resulting iterative approaches is analyzed from a theoretical point of view. Moreover, we propose an agenda for further research in this field.     

Robust real-time pedestrians detection in urban environments with low-resolution cameras

Detecting that pedestrians are present in front of a vehicle is highly desirable to avoid dangerous traffic situations. A novel vision-based system is presented to automatically detect far-away pedestrians with low-resolution cameras mounted in vehicles given the contributions of fixed cameras present in the scene.Fixed cameras detect pedestrians by solving an inverse problem built upon a multi-class dictionary of atoms approximating the foreground silhouettes. A sparse-sensing strategy is proposed to extract the foreground silhouettes and classify them in real-time. Mobile cameras detect pedestrians given only their appearance in the fixed cameras. A cascade of compact binary strings is presented to model the appearance of pedestrians and match them across cameras.The proposed system addresses the practical requirements of transportation systems: it runs in real-time with low memory loads and bandwidth consumption. We evaluate the performance of our system when extracted features are severely degraded and the sensing devices are of low quality. Experimental results demonstrate the feasibility of our collaborative vision-based system.     

Connected cruise control among human-driven vehicles: Experiment-based parameter estimation and optimal control design

In this paper, we consider connected cruise control design in mixed traffic flow where most vehicles are human-driven. We first propose a sweeping least square method to estimate in real time feedback gains and driver reaction time of human-driven vehicles around the connected automated vehicle. Then we propose an optimal connected cruise controller based on the mean dynamics of human driving behavior. We test the performance of both the estimation algorithm and the connected cruise control algorithm using experimental data. We demonstrate that by combining the proposed estimation algorithm and the optimal controller, the connected automated vehicle has significantly improved performance compared to a human-driven vehicle.     

A dynamic automated lane change maneuver based on vehicle-to-vehicle communication

Automated driving is gaining increasing amounts of attention from both industry and academic communities because it is regarded as the most promising technology for improving road safety in the future. The ability to make an automated lane change is one of the most important parts of automated driving. However, there has been little research into automated lane change maneuvers, and current research has not identified a way to avoid potential collisions during lane changes, which result from the state variations of the other vehicles. One important reason is that the lane change vehicle cannot acquire accurate information regarding the other vehicles, especially the vehicles in the adjacent lane. However, vehicle-to-vehicle communication has the advantage of providing more information, and this information is more accurate than that obtained from other sensors, such as radars and lasers. Therefore, we propose a dynamic automated lane change maneuver based on vehicle-to-vehicle communication to accomplish an automated lane change and eliminate potential collisions during the lane change process. The key technologies for this maneuver are trajectory planning and trajectory tracking. Trajectory planning calculates a reference trajectory satisfying the demands of safety, comfort and traffic efficiency and updates it to avoid potential collisions until the lane change is complete. The trajectory planning method converts the planning problem into a constrained optimization problem using the lane change time and distance. This method is capable of planning a reference trajectory for a normal lane change, an emergency lane change and a change back to the original lane. A trajectory-tracking controller based on sliding mode control calculates the control inputs to make the host vehicle travel along the reference trajectory. Finally, simulations and experiments using a driving simulator are conducted. They demonstrate that the proposed dynamic automated lane change maneuver can avoid potential collisions during the lane change process effectively.     

A real-time system for monitoring driver fatigue

ABSTRACT

This paper presents a nonintrusive prototype computer vision system for real-time fatigue driving detection. First, we use Haar-like features to detect a driver’s face and conduct tracking by introducing an improved Camshift algorithm. Second, we propose a new eye-detection algorithm that combines the Adaboost algorithm with template matching to reduce computational costs and add an eye-validation process to increase the accuracy of the detection rate. Third, and different from other methods focusing on detecting eyes using the ‘bright pupil’ effect, which only works well only for certain constrained lighting conditions, our method detects and estimates the iris center in the hue (H) channel of the hue, saturation, value color space and fits the iris with an ellipse. After extracting the eye fatigue features, we calculate the PERCLOS measurement for fatigue evaluation. This system has been tested on the IMM Face Database, which contains more than 200 faces, and in a real-time test. The experimental results show that the system possesses good accuracy and robustness.     

Sensor location problems in path-differentiated congestion pricing

Path-differentiated congestion pricing is a tolling scheme that imposes tolls on paths instead of individual links. One way to implement this scheme is to deploy automated vehicle identification sensors, such as toll tag readers or license plate scanners, on roads in a network. These sensors collect vehicles’ location information to identify their paths and charge them accordingly. In this paper, we investigate how to optimally locate these sensors for the purpose of implementing path-differentiated pricing. We consider three relevant problems. The first is to locate a minimum number of sensors to implement a given path-differentiated scheme. The second is to design an optimal path-differentiated pricing scheme for a given set of sensors. The last problem is to find a path differentiated scheme to induce a given target link-flow distribution while requiring a minimum number of sensors.     

A driving cycle detection approach using map service API

Following advancements in smartphone and portable global positioning system (GPS) data collection, wearable GPS data have realized extensive use in transportation surveys and studies. The task of detecting driving cycles (driving or car-mode trajectory segments) from wearable GPS data has been the subject of much research. Specifically, distinguishing driving cycles from other motorized trips (such as taking a bus) is the main research problem in this paper. Many mode detection methods only focus on raw GPS speed data while some studies apply additional information, such as geographic information system (GIS) data, to obtain better detection performance. Procuring and maintaining dedicated road GIS data are costly and not trivial, whereas the technical maturity and broad use of map service application program interface (API) queries offers opportunities for mode detection tasks. The proposed driving cycle detection method takes advantage of map service APIs to obtain high-quality car-mode API route information and uses a trajectory segmentation algorithm to find the best-matched API route. The car-mode API route data combined with the actual route information, including the actual mode information, are used to train a logistic regression machine learning model, which estimates car modes and non-car modes with probability rates. The experimental results show promise for the proposed method’s ability to detect vehicle mode accurately.     

On the network connectivity of platoon-based vehicular cyber-physical systems

In the past few years, vehicular ad hoc networking (VANET) has attracted significant attention and many fundamental issues have been investigated, such as network connectivity, medium access control (MAC) mechanism, routing protocol, and quality of service (QoS). Nevertheless, most related work has been based on simplified assumptions on the underlying vehicle traffic dynamics, which has a tight interaction with VANET in practice. In this paper, we try to investigate VANET performance from the vehicular cyber-physical system (VCPS) perspective. Specifically, we consider VANET connectivity of platoon-based VCPSs where all vehicles drive in platoon-based patterns, which facilitate better traffic performance as well as information services. We first propose a novel architecture for platoon-based VCPSs, then we derive the vehicle distribution under platoon-based driving patterns on a highway. Based on the results, we further investigate inter-platoon connectivity in a bi-directional highway scenario and evaluate the expected time of safety message delivery among platoons, taking into account the effects of system parameters, such as traffic flow, velocity, platoon size and transmission range. Extensive simulations are conducted which validate the accuracy of our analysis. This study will be helpful to understand the behavior of VCPSs, and will be helpful to improve vehicle platoon design and deployment.     

Development of a sensor system for traffic data collection

Although many types of traffic sensors are currently in use, all have some drawbacks, and widespread deployment of such sensor systems has been difficult due to high costs. Due to these deficiencies, there is a need to design and evaluate a low cost sensor system that measures both vehicle speed and counts. Fulfilling this need is the primary objective of this research. Compared to the many existing infrared-based concepts that have been developed for traffic data collection, the proposed method uses a transmission-based type of optical sensor rather than a reflection-based type. Vehicles passing between sensors block transmission of the infrared signal, thus indicating the presence of a vehicle. Vehicle speeds are then determined using the known distance between multiple pairs of sensors. A prototype of the sensor system, which uses laser diode and photo detector pairs with the laser directly projected onto the photo detector, was first developed and tested in the laboratory. Subsequently this experimental prototype was implemented for field testing. The traffic flow data collected were compared to manually collected vehicle speed and traffic counts and a statistical analysis was done to evaluate the accuracy of the sensor system. The analysis found no significant difference between the data generated by the sensor system and the data collected manually at a 95% confidence interval. However, the testing scenarios were limited and so further analysis is necessary to determine the applicability in more congested urban areas. The proposed sensor system, with its simple technology and low cost, will be suitable for saturated deployment to form a densely distributed sensor network and can provide unique support for efficient traffic incident management. Additionally, because it may be quickly installed in the field without the need of elaborate fixtures, it may be deployed for use in temporary traffic management applications such as traffic management in road work zones or during special events.     

Driving analytics using smartphones: Algorithms,comparisons and challenges

The present work investigates the use of smartphones as an alternative to gather data for driving behavior analysis. The proposed approach incorporates i. a device reorientation algorithm, which leverages gyroscope, accelerometer and GPS information, to correct the raw accelerometer data, and ii. a machine-learning framework based on rough set theory to identify rules and detect critical patterns solely based on the corrected accelerometer data. To evaluate the proposed framework, a series of driving experiments are conducted in both controlled and “free-driving” conditions. In all experiments, the smartphone can be freely positioned inside the subject vehicle. Findings indicate that the smartphone-based algorithms may accurately detect four distinct patterns (braking, acceleration, left cornering and right cornering) with an average accuracy comparable to other popular detection approaches based on data collected using a fixed position device.     

A novel physiological index for Driver’s Activation State derived from simulated monotonous driving studies

Automobile driving in monotonous situations such as driving for long periods and/or travelling a familiar route may cause the lowering of the driver’s awareness level or what we term here as a Driver’s Activation State (DAS), resulting in an increased risk of an accident. We propose here to develop means with which to create an in-car environment so as to allow active driving, hopefully thus avoiding potentially dangerous situations. In order ultimately to develop a validated activation method, we firstly set out to examine physiological variables, including cardiovascular parameters, during simulated monotonous driving. Subsequently, we investigated the derivation of a suitable DAS index. During the experiment, a momentary electrical test stimulus of 0.5 s duration was applied at a rate of approximately once per 10 min to the subject’s shoulder to evoke a physiological responses. In 11 healthy male volunteers we successfully monitored physiological variables during the experiment and found particular patterns in the beat-by-beat changes of blood pressure in response to the electrical test stimulus. This finding, explained by autonomic activity balance, suggests that the patterns may be used as an appropriate and practicable index relevant to the Driver’s Activation State.     

Rapid estimation of electric vehicle acceptance using a general description of driving patterns

A reliable estimate of the potential for electrification of personal automobiles in a given region is dependent on detailed understanding of vehicle usage in that region. While broad measures of driving behavior, such as annual miles traveled or the ensemble distribution of daily travel distances are widely available, they cannot be predictors of the range needs or fuel-saving potential that influence an individual purchase decision. Studies that record details of individual vehicle usage over a sufficient time period are available for only a few regions in the US. In this paper we compare statistical characterization of four such studies (three in the US, one in Germany) and find remarkable similarities between them, and that they can be described quite accurately by properly chosen set of distributions. This commonality gives high confidence that ensemble data can be used to predict the spectrum of usage and acceptance of alternative vehicles in general. This generalized representation of vehicle usage may also be a powerful tool in estimating real-world fuel consumption and emissions.