Search Results

This paper describes a human factors simulation study of the decision making behavior of drivers attempting to avoid nonrecurring congestion by diverting to alternate routes with the aid of in-vehicle navigation systems. This study is the first phase of a two part project in which the second phase will apply the driver behavior data to a simulation model analysis of traffic flow. The object of the driver behavior experiment was to compare the effect of various experimental navigation systems on driver route diversion and alternate route selection. The experimental navigation system configurations included three map based systems with varying amounts of situation information and a non map based route guidance system. The overall study results indicated that navigation system characteristics can have a significant effect on driver diversion behavior, with better systems allowing more anticipation of traffic congestion.

A tire/terrain interaction model is presented to support the dynamic simulation of off-road ground vehicle. The model adopts a semi-empirical approach that is based on curve fits of soil data combined with soil mechanics theories to capture soil compaction, soil shear deformation, and soil passive failure that associate with off-road driving. The resulting model allows the computation of the tire forces caused by terrain deformation in longitudinal and lateral direction. This model has been compared with experimental data and shown reasonable prediction of the tire/terrain interaction.

Basic performance properties of tires significantly influence the lateral/directional (steering) stability and handling of highway vehicles. These properties include cornering stiffness and peak and slide coefficients of friction. This paper considers some detailed tire machine measurements of lateral tire performance. A large database of tire properties for a wide range of highway vehicles is also analyzed. A regression analysis approach is used to define the sensitivity of various size and operational (speed, pressure and load) characteristics on tire behavior. The paper discusses the manner in which these properties vary with tire size and operational conditions, and the effect of the properties on vehicle stability and handling.

This paper describes the driver/vehicle modeling aspects of a computer simulation that can respond to highway engineering descriptions of roadways. The driver model interacts with a complete vehicle dynamics model that has been described previously. The roadway path is described in terms of horizontal and vertical curvature and cross slopes of lanes, shoulders, side slopes and ditches. Terrain queries are made by the vehicle dynamics to locate tires on the roadway cross-section, and to define vehicle path and road curvature at some distance down the road. The driver model controls steering to maintain lateral lane position. Speed is maintained at a speed limit on tangents, and decreased as needed to maintain safe lateral acceleration. Because the bandwidth of longitudinal (speed) control is much lower than lateral/directional (steering) control, the driver model looks further ahead for speed control than for steering.

This paper describes a laboratory simulation study of driver reaction to in-vehicle navigation systems. The study included a pre-test questionnaire on demographic background and commuting behavior, simulation testing of navigation decision making, and a post-test questionnaire on navigation behavior and reactions to in-vehicle navigation systems and the laboratory simulation. A total of 277 subjects, both male and female, were employed over a wide range of ages. Test subjects were assigned to one of four navigation system groups or a no-system control group for the purpose of comparing system performance. The simulation task required subjects to experience a commuting ‘drive’ on a Southern California freeway route and minimize trip time by diverting off the main route to avoid congestion. Subjects were given orientation and training on the simulation and their navigation system condition, and were motivated by rewards and penalties to minimize trip time.

Motion cueing algorithms can improve the perceived realism of a driving simulator, however, data on the effects on driver performance and simulator sickness remain scarce. Two novel motion cueing algorithms varying in concept and complexity were developed for a limited maneuvering workspace, hexapod/Stuart type motion platform. The RideCue algorithm uses a simple swing motion concept while OverTilt Track algorithm uses optimal pre-positioning to account for maneuver characteristics for coordinating tilt adjustments. An experiment was conducted on the US Army Tank Automotive Research, Development and Engineering Center (TARDEC) Ride Motion Simulator (RMS) platform comparing the two novel motion cueing algorithms to a pre-existing algorithm and a no-motion condition.

Real time man-in-the-loop simulation can be used in a variety of research, testing and training roles where safety, efficiency and/or economy are important. Simulation can allow complete control and uniformity over driving conditions and permit analysis of a range of vehicle and driver behavior variables. Simulation complexity and fidelity requirements will vary depending on application requirements. This paper reviews past and current driving simulation development efforts and applications. Simulation requirements are assessed relative to various applications, including vehicle handling, driver behavior, training, licensing and fitness for duty testing.

Interactive simulation requires providing appropriate sensory cuing and stimulus/response dynamics to the driver. Sensory feedback can include visual, auditory, motion, and proprioceptive cues. Stimulus/response dynamics involve reactions of the feedback cuing to driver control inputs including steering, throttle and brakes. The stimulus/response dynamics include both simulated vehicle dynamics, and the response dynamics of the simulation hardware including computer processing delays. Typically, simulation realism will increase with sensory fidelity and stimulus/response dynamics that are equivalent to real-world conditions (i.e. without excessive time delay or phase lag). This paper discusses requirements for sensory cuing and stimulus/response dynamics in real-time, interactive driving simulation, and describes a modest fixed-base (i.e. no motion) device designed with these considerations in mind.

Interactive driving simulation is attractive for a variety of applications, including screening, training and licensing, due to considerations of safety, control and repeatability. However, widespread dissemination of these applications will require modest cost simulator systems. Low cost simulation is possible given the application of PC level technology, which is capable of providing reasonable fidelity in visual, auditory and control feel cuing. This paper describes a PC based simulation with high fidelity vehicle dynamics, which provides an easily programmable visual data base and performance measurement system, and good fidelity auditory and steering torque feel cuing. This simulation has been used in a variety of applications including screening truck drivers for the effects of fatigue, research on real time monitoring for driver drowsiness and measurement of the interference effect of in-vehicle IVHS tasks on driving performance.

A combined biodynamic and vehicle model is used to assess the vibration and performance of a human operator performing driving and other tasks. The other tasks include reaching, pointing and tracking by the driver and/or passenger. This analysis requires the coordinated use of separate and mature software programs for anthropometrics, vehicle dynamics, biodynamics, and systems analysis. The total package is called AVB-DYN, an acronym for Anthropometrics, Vehicle, and Bio-DYNamics. The objectives and architecture are discussed, and then a preliminary version of this package is demonstrated in an example where a HMMWV (High Mobility Multipurpose Wheeled Vehicle) operator is performing a driving task.

A combined biodynamic and vehicle model is used to assess the vibration and performance of a human operator performing driving and other tasks. The other tasks include reaching, pointing and tracking by the driver and/or passenger. This analysis requires the coordinated use of separate and mature software programs for anthropometrics, vehicle dynamics, biodynamics, and systems analysis. The total package is called AVB-DYN, an acronym for Anthropometrics, Vehicle and Bio-DYNamics. The biodynamic component of AVB-DYN is described, and then compared with an experiment that studied human operator in-vehicle reaching performance using the U.S. Army TACOM Ride Motion Simulator.

The U.S. Army uses the root mean square and power spectral density of elevation to characterize road/terrain (off-road) roughness for durability. This paper describes research aimed toward improving these metrics. The focus is on taking previously developed metrics and applying them to mathematically generated terrains to determine how each metric discerns the relative roughness of the terrains from a vehicle durability perspective. Multiple terrains for each roughness level were evaluated to determine the variability for each terrain rating metric. One method currently under consideration is running a relatively simple, yet vehicle class specific, model over a given terrain and using predicted vehicle response(s) to classify or characterize the terrain.

The U.S. Army uses the root mean square and power spectral density of elevation to characterize road/terrain (off-road) roughness for durability. This paper describes research aimed toward improving these metrics. The focus is on taking previously developed metrics and applying them to mathematically generated terrains to determine how each metric discerns the relative roughness of the terrains from a vehicle durability perspective. Multiple terrains for each roughness level were evaluated to determine the variability for each terrain rating metric. One method currently under consideration is running a relatively simple, yet vehicle class specific, model over a given terrain and using predicted vehicle response(s) to classify or characterize the terrain.

A Grade Severity Rating System (GSRS) was developed as a means for reducing the incidence and severity of truck accidents on downgrades. The ultimate result is a roadside sign at the top of each hill. The sign is tailored to the individual hill and gives a recommended maximum speed (to be held constant for the entire grade descent) for each of several truck weight ranges. This concept represents a major step forward in terms of grade descent safety because it tells the driver what to do directly, rather than giving him information which still requires evaluation under different loading conditions.

Driver response and performance can be quantified by observing the stimulus-response environment. Yet the driver's inherent adaptability allows him to have seemingly adequate performance in potentially hazardous driving situations even though he may be operating near the acceptable safety limits. Physiological measures of the driver's internal state can provide further quantification of his performance level and can give a measure of his workload or safety performance margin. Measures of driver physiological and control responses have been made under gust disturbance conditions with the subject's car operating at various speeds. The experimental techniques and data are described, and correlations between the situational parameters and driver stress and control response are shown.

This paper reviews and expands previously published driver steering control models. The driver model is structured to control vehicle heading angle and lane position. Field test data are used to validate model structure. The closed-loop stability of the driver/vehicle system is analyzed using a two degree of freedom vehicle dynamics approximation. This analysis is used to develop constraints among the various driver model parameters and their dependence on vehicle characteristics. Driver/vehicle model approximations are also used to explore the effects of driver behavior on steering performance.

The driver's ability to control the lateral position of an automobile is dependent on his perception of the command path (roadway) to be followed. This perception is affected by both the configuration of road markings and other features, and the visibility of these elements. As visibility decreases, the driver's preview of the commanded path is reduced. Theory indicates that driver performance should degrade with reduced preview and configurational parameters which characterize the intermittent nature of delineation (e.g., dashed lines). This paper describes a simulation experiment in which driver behavior and driver/vehicle system performance were measured over a range of visibility and configuration parameter variations. Driver dynamic response and noise (remnant) were reliably affected by variations in visibility and configuration. These effects were also reflected in system performance measures such as lane deviations.

For almost twenty years, researchers have attempted to develop an in-vehicle system which would prevent an impaired driver from operating his or her motor vehicle. These systems have ranged from breath testers to psychomotor tests, and have prevented operation of the vehicle by such methods as preventing the vehicle from starting or alerting drivers, and the police through alarm systems. This paper discusses the background leading to an in-vehicle system which was built and tested. We also discuss the system and its components, and present the results of two tests involving convicted drunk drivers. While the primary purpose of this project was to determine the feasibility of this type of system, the results of the two tests show promise for the reduction of impaired driving trips.

This paper presents an analysis of driver/vehicle performance over a range of maneuvering conditions including accident avoidance scenarios involving vehicle limit performance handling. Driver behavior is considered in the same dynamic analysis terms as vehicle response in order to give appropriate closed-loop measures of total system maneuvering capability and handling stability. A driver control structure is developed along with closed-loop system stability constraints on model parameters over a wide range of vehicle maneuvering conditions. Example simulation runs are presented for several accident avoidance scenarios.

This paper describes the results of an experiment concerning driver behavior in car following tasks. The motivation for this experiment was a desire to understand typical driver car following behavior as a guide for setting the automatic control characteristics of an ACC (Adaptive Cruise Control) system. Testing was conducted under both test track and open road driving conditions. The results indicate that car following is carried out under much lower bandwidth conditions than typical steering processes. Dynamic analysis shows driver time delay in response to lead vehicle velocity change on the order of several seconds. Typical longitudinal acceleration distributions show standard deviations of less than 0.05 g (acceleration due to gravity).