Vehicle Dynamics

  •    Active Projects

       Select Previous Projects

    • Do you think this research is awesome? We do too! If you really like our research and would like to join the group, please read the following document:

    • Potential group members.docx
  • Detecting the Instability of Oncoming Vehicles Using Optical Flow and Map-Based Context

    Investigators: Chris Monaco, Dr. Sean Brennan

    This research seeks to determine the feasibility of detecting the instability of oncoming vehicles using optical flow from a stereo camera. A "visual odometry" method was selected to estimate the egomotion of an instrumented mapping vehicle from its onboard stereo camera and inertial measurement unit (IMU) data. This method analyzes optical flow to estimate the egovehicle's translational and rotational motion. Next, the algorithm classifies features as either static or dynamic to detect dynamic vehicles. Once detected, optical flow analysis then estimates the external vehicle's states using extracted features from the moving egovehicle. This process enables the novel concept of using optical flow to detect an oncoming vehicle's instability.

    The detection of vehicle instability is greatly aided by location context that can be used to warn drivers of imminent risk. To provide context, this research considers the use of a map to isolate instabilities that may result in entry of the egovehicle's lane. Specifically, a map provides additional knowledge of an external vehicle's road radius from position measurements and map data. This permits the estimation of a external vehicle's neutral steering yaw rate. If the external vehicle's measured yaw rate is in excess of its estimated neutral steering yaw rate, this then indicates instability and a likelihood that the vehicle will be unable to maintain consistent lane tracking.

    The underlying algorithms were tested offline with data collected from The Pennsylvania State University's mapping vehicle and a precisely controlled test vehicle at The Pennsylvania State University Larson Transportation Institute Test Track. While this concept's validation was limited - and not the direct goal of this research - the results show promise for this concept's feasibility and future validation.

    Figure 1. Top: Annotated video of the algorithm estimating the egovehicle's velocities in 6DoF (green), detecting an external dynamic vehicle, and estimating the external dynamic vehicle's location, orientation, and velocities (red). Bottom: Screenshot of video.

    This work is intended to be implemented as part of a more robust and complex real-time architecture. When deployed as an Advanced Driver Assistance System (ADAS) on a moving vehicle, it can ensure early detection of an oncoming vehicle's instability, prompting increased caution to mitigate collision risks. Furthermore, when deployed statically at an intersection, it can detect low friction-areas based on common areas of vehicle instability, prompting remedial action to fix the roadway. In both situations, this novel concept has the potential to reduce roadway fatalities and serious injuries.

  • Development of an Open-Source Tractor Simulator and Hardware-In-The-Loop Environment based on the Robot Operating System and the Blender Game Engine

    Investigators: Nicolas Ochoa, Dr. Sean Brennan, Dr. H.J. Sommer III, Dr. Dennis Murphy
    Former Investigators: Kevin Swanson, Dr. Alexander Brown

    The goal of this project is to study how vehicle operators interact with technological assistance systems in safety-critical situations. For this purpose, the vehicle dynamics simulator was developed using open source software—the Robot Operating System (ROS) to interface with hardware, plus the Blender Game Engine to generate the 3D virtual environments and simulations— a MOOG motion base with 6 degrees of freedom and a 360 degree surround screen.

    Current research is focused on preventing rollover incidents on farm tractors by addressing two key issues. First, studying how the driver’s perceptual errors of roll and pitch angles might affect their assessment of rollover risk. Second, determining how an alert system could be used to improve the operator’s awareness while driving in high rollover risk situations.

    The simulator’s modular software allows it to change virtual environments and vehicle dynamics, as well as connect to other Hardware-in-the-loop platforms—including other driving simulators.

  • Vehicle Road Departure Detection Using Anomalies In Dynamics

    Investigators: Hang Yang, Dr. Sean Brennan

    Roadway departure is a leading cause of vehicle accidents, and most recent research on roadway departure prevention systems focuses on on-pavement trajectory recovery. However, the detection of lane departures is not easy, and current methods are heavily dependent on vision-based sensors. Unfortunately, such lane detection often does not work under inclement weather conditions, which are a major cause of roadway departure accidents. Few studies have investigated whether one can detect road departure by using other sensors, for example detecting changes in tire forces and resulting body motion due to transitions to off-pavement conditions.

    This research aims to identify transition from on-road to off-road condition by looking at changes in vehicle dynamics, specifically in vehicle yaw rate and sideslip angle. This approach is based on well-known vehicle models and comparing these to measurements from a GPS/INS system to obtain real-time estimates of model agreement with measured data.

  • Lateral Vehicle State and Environment Estimation Using Temporally Previewed Mapped Lane Features

    Investigators: Alexander Brown, Dr. Sean Brennan

    This project investigates a model-based method for using a forward-looking monocular camera along with previewed road geometry from a high-fidelity, low-dimensional map to estimate lateral planar vehicle states by measuring the vehicle's temporally anticipated reference trajectory. It seeks to combine a computationally cheap, simple, and extensible feature detection strategy with the benefits of map-based estimation within a relatively simple estimation framework that gives a-priori predictions of algorithm performance. This is accomplished using a planar, linear "bicycle model" for lateral vehicle dynamics, which is shown to be sufficiently accurate for normal driving conditions. Rather than attempting to produce the best lane detector or state estimation strategy possible, this approach seeks to produce a framework that helps perception system designers determine the minimum amount of lookahead distance needed to achieve the best lateral vehicle state estimates possible, especially with low-cost sensors.

  • Temporal Preview Estimation for Design of an Optimal Preview Vehicle Guidance System Using a Foward-Facing Monocular Camera

    Investigators: Alexander Brown, Dr. Sean Brennan

    This project investigates using a forward- looking monocular camera along with previewed road geometry from a high-fidelity, low-dimensional map to estimate lateral planar vehicle states and guide a vehicle down a road using optimal preview control with estimated system states. Concrete performance and robustness metrics for the map-based stochastic preview control/estimation architecture, derived from linear system theory, allow for the design of control and perception systems maintain safety given a calculable minimum visibility threshold.

  • Preview Horizon Analysis for Vehicle Rollover Prevention

    Investigators: Paul Stankiewicz, Dr. Alexander Brown, Dr. Sean Brennan

    Vehicle rollover remains one of the deadliest types of automobile accident. This research estimates the minimum preview time needed to prevent untripped wheel lift events by analyzing simple maneuvers expected in collision avoidance situations. To predict a vehicle's future rollover propensity, the Zero-Moment Point (ZMP) metric is applied to projected vehicle trajectories. Comparing different amounts of preview, the results show that short-range predictions - as little as 0.7 seconds ahead of the vehicle - are sufficient to prevent nearly all dynamics-induced rollovers in typical highway curves. These results are useful to determine the minimum preview windows that may be necessary for more advanced vehicle control methods, such as Model Predictive Control.

  • Development of an Open-Source Vehicle Dynamics Simulator and Hardware-In-The-Loop Environment based on the Robot Operating System (ROS) and the Gazebo Simulation Environment

    Investigators: Nicolas Ochoa, Kevin Swanson, Dr. Alexander Brown, Dr. Sean Brennan

    This project deals with the development of a Hardware-in-the-Loop (HIL) driving simulator designed with ROS and Gazebo. Analyses of current driving simulator software has unearthed a need for a simulation environment that also functions as a data acquisition and control framework, inclusive of HIL test functionality. Work is ongoing to develop an open-source HIL-centric driving simulator, with a close eye on optimizing scalable system performance in terms of system latency and extensibility.

  • Terrain-Aware Rollover Prediction for Ground Vehicles

    Investigators: Sittikorn Lapapong, Alexander Allen Brown Dr. Sean Brennan

    Rollover accidents are one of the leading causes of death on highways. A key challenge in preventing rollover via chassis control is the accurate prediction of the onset of rollover. This task is especially hard in the presence of terrain features typical of roadway environments, such as road superelevation (i.e. road bank), the median slope, and the shoulder down-slope. This project develops a vehicle rollover prediction algorithm that is based on a kinematic analysis of vehicle motion, a method that allows explicit inclusion of terrain effects. The solution approach utilizes the concept of Zero-Moment Point (ZMP) that is typically applied to walking robot dynamics. This concept is introduced in terms of a lower-order model of vehicle roll dynamics to measure the vehicle rollover propensity, and the resulting ZMP prediction allows a direct measure of a vehicle rollover threat index.

    PREDICTING VEHICLE ROLLOVER: Experimental maneuvers with unmanned vehicles are used in testing the accuracy of rollover prediction algorithms.

  • Dynamics and Dimensional Similitude of Vehicle Impacts on Soil-Fixed Boulders

    Investigators: Mark Keske, Sam Hoskins, Dr. Alexander Brown, Dr. Sean Brennan

    This project focuses on the development and validation of a low-order model for a vehicle impact on a soil-fixed boulder embedded in cohesionless soil. Dimensional analysis is then applied to the low-order model to develop dimensionless equations of motion and scaling laws associated with small and full scale testing. Additionally, in-situ soil measurement methods are proposed which would enable rapid determination of the soil properties found within the low-order model, such as soil density and modulus of subgrade reaction.

    The vehicle is modeled as a lumped-parameter Maxwell model, the boulder is treated as a rigid body, and the soil is modeled as a system of lumped-parameter Kelvin models. The Maxwell model parameters for the vehicle are found by correlating an expected displacement response for a vehicle impacting a rigid wall to measured data from finite element simulations. The Kelvin model parameters for the soil are found through the proposed methods of in-situ soil testing and a derivation of soil damping using conservation of momentum.

    The measured and estimated lumped-parameter values are then used to simulate a vehicle impact on a soil-fixed boulder using numerical integration techniques. The simulation is then compared and validated against past full scale crash tests. The dimensionless equations of motion and scaling laws are used to simulate small scale vehicle impacts as well as correlate measured displacements from small scale crash tests to full scale crash tests.

    VEHICLE-BOULDER COLLISIONS: Vehicle collision tests with boulders of various dimensions are carried out to validate the results from analytical studies and numerical simulations.

  • Design of Horizontal Highway Curves with Downgrades

    Investigators: Tejas Varunjikar, Dr. Sean Brennan

    Geometric design of highways is an important aspect of highway engineering, and in particular, horizontal curves on highways have higher accident rates compared to straight roads. Quantitative guidelines for horizontal curve design exist only for flat roads, but not downgrades. This project focuses on developing suitable vehicle dynamics models to determine acceptable horizontal curve geometries on downgrades. A friction demand versus friction supply approach is used to check whether the current horizontal curve design policies are acceptable for downgrades. Skid measurements from the field combined with a physics-based tire model are used to obtain the friction supply at various design speeds. This project develops analytical as well as low-order simulation-based models like bicycle model for a vehicle traveling on downgrade in order to find the friction demand of the vehicle. Results show that per-axle friction demand can be significantly higher compared to the overall friction demand which is basis of current design guidelines. The margins of safety are shown to significantly decrease with design speed, and in the case of even moderate braking, go to a very low value at high speeds.

  • Select Previous Projects

  • Highway Median Safety Analysis through Vehicle Dynamics Simulations

    Investigators: Jason Stine, Dr. Sean Brennan, Dr. Eric Donnell
    Sponsors: NCHRP 22-21 pooled fund in partnership with the Midwest Research Institute, The Thomas D. Larson Pennsylvania Transportation Institute (LTI), and the Mid-Atlantic Universities Transportation Center (MAUTC)

    In highway design, there is, in general, little consideration of the vehicle dynamics in development of highway and road-side features. The few software packages available to combine highway design with vehicle trajectory predictions as a rule do not consider driver reaction to roadside features, nor allow for driver’s steering input. Even though simulations have been used to aid in some highway design applications, the use of multi-body simulations to evaluate roadway designs remains rare.

    In contrast, for decades vehicle dynamics simulation packages have been used for evaluation of vehicle performance, stability, and accident reconstruction. But these studies in general assume a flat roadway surface, so the influence of terrain on vehicle behavior is essentially negligible.

    With the increasing number of in-median rollovers, and increasing concern for cross-median frontal impacts, there is growing demand for improved median and shoulder designs to prevent the thousands of deaths per year caused by vehicle interaction with median and shoulder slopes. This study, using CarSim software, aims at evaluating the safety of highway medians through the simulation of off-road excursions. Several different vehicle classes, speeds, encroachment angles, and driver inputs were tested in the simulations. Resulting vehicle positions and state are being analyzed for various median profiles, with emphasis on investigating the design tradeoff between preventing vehicles from rolling over or entering the opposing lanes of traffic. The results from this study will support future highway safety evaluation, as well as the design and placement for new in-median cable barriers.

    CARSIM SIMULATION:The simulation depicts the influence of the driver's input. Also shown here are the two catastrophic results seen in median incursions; vehicle rollover (white) and entrance into opposing lane (yellow).

  • Immersive Driving Simulator Interface with CarSimDS Live Vehicle Dynamics Software

    Investigators: Jason Stine, Cynthia LaJambe (LTI), Dr. Sean Brennan
    Sponsors: The Thomas D. Larson Pennsylvania Transportation Institute

    To study the interaction of the human driver with a vehicle, this project has begun rebuilding the existing driving simulator at the Thomas D. Larson Pennsylvania Transportation Institute with the live driving simulator software provided by CarSim. Current work incorporates the interfacing of three live animators from CarSim with three frontal projection screens. This will provide a 140 degree field of vision from the driver seat of the Mac truck cab. Future work will involve combining this system with EEG and eye-tracker measurements, as well as interfacing the driving simulator with live, external vehicles to allow for realistic study of human-driver interaction.

  • Pursuit Management: Attribute-based Evaluation of Stop stick, Spike strip and Other Tire Deflation Technologies

    Investigators: Alex Brown, Dr. Sean Brennan, Dr. Zoltan Rado, Dr. Robert Gray
    Sponsors: National Institute of Justice

    There are many high-speed pursuit situations where there are simply no technological means to stop a pursued vehicle, and there are also many situations known to be quite dangerous to the officers involved. The National Institute of Justice, in a long-term effort to improve the effectiveness and safety of police pursuits, wishes to evaluate the safety and effectiveness issues involved with terminating a pursuit using a tire deflation device to incapacitate a pursued vehicle. This study will involve testing the commonly used devices at Mid-State Airport at varying speeds, varying tire technology, and with/without concurrent vehicle maneuvers i.e. a swerve. The test vehicle will be a fully instrumented, fully automated (un-manned) LTI fleet vehicle configured to emulate a car that may typically be involved in a high-speed police pursuit.

  • Distributions of Vehicle Parameters and their associated Dimensionless Parameters

    Investigators: Sittikorn Lapapong, Dr. Sean Brennan

    This study involves analyzing the distributions of vehicle parameters and their associated dimensionless parameters. The vehicle parameters used in the study are related to the bicycle model as well as roll and pitch dynamics. The parameters have been primarily collected and synthesized from the database of the National Highway Traffic Safety Administration (NHTSA) and more than 270 vehicle-dynamics-related literatures. A group of non-dimensional parameters has been created using dimensional analysis and the Buckingham Pi theorem. The reduced number of dimensionless parameters is being used to analyze system similarity across size scales.

    DISTRIBUTION OF VEHICLE MASSES: The vehicle mass distribution from the NHTSA database resembles that from the literature, indicating good agreement between the two data sets.