Pith. sign in

REVIEW 3 major objections 5 minor 52 references

Beyond the Plane: Coupling Planar Vehicle Dynamics with Three-Dimensional Road Geometry

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A wrapping module lets a planar vehicle dynamics model reproduce measured three-dimensional accelerations and yaw rates on banked roads, cutting mean absolute error by roughly 95% for vertical and 78% for lateral acceleration.

desk verdict A useful black-box module for adding 3D road effects to planar vehicle models, but the headline real-data numbers validate the kinematic transformation, not the force-feedback dynamics. read the letter →

arxiv 2608.09402 v1 pith:QMAFTVS3 submitted 2026-08-10 cs.RO

Simon Sagmeister , Phillip Pitschi , Nico Haja , Markus Lienkamp This is my paper · ORCID
classification cs.RO
keywords 3Droadgeometryplanarvehicledynamicsribbontrackrepresentationbankedvalidationautonomousracingsimulationroad-geometryforcefeedbackreal-timecurvilinearcoordinates
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Most vehicle dynamics simulations treat the road as flat, but real roads are banked and sloped, and on the Las Vegas Motor Speedway the 3D geometry raises tire normal loads by more than 66% over standstill. This paper claims that a planar vehicle model can nevertheless reproduce measured 3D behavior if it is coupled to the road through a ribbon representation: project the 3D road onto a plane that preserves curvature and arc length, drive the planar model on that flattened road, transform its outputs back to 3D, and feed the geometry-induced forces and moments back at the center of gravity. Validated against full-scale race-car data at 69.5 m/s, the method reduces mean absolute error versus the planar baseline by about 95% for vertical acceleration, 78% for lateral acceleration, and 82% for mean yaw-rate error, at a mean cost of 9.5 microseconds per simulation step. The claim matters because it offers a path to accurate 3D-road simulation without abandoning interpretable planar models or paying multi-body complexity.

What carries the argument

The load-bearing object is the ribbon track representation: a 3D road described by a spine curve plus a lateral offset, with curvature $\Omega_z$, banking $\varphi$, and slope $\mu$ that depend only on arc length $s$. Onto this ribbon the method projects a 'road plane'—the surface cut into infinitesimal slices and laid flat—constructed by numerically integrating the curvature with a circular-arc scheme that avoids accumulation error at low curvature. Curvilinear coordinates $(s, n, \chi)$ are preserved between the plane and the 3D ribbon, which makes the pose transformation well-defined, with the full orientation built from $R_{IV} = R_z(\theta)R_y(\mu)R_x(\varphi)R_z(\chi-\beta)$. The measurement transformations (velocity, angular velocity, acceleration, angular acceleration) and the vehicle-load calculation ($F = ma$ plus angular momentum with a diagonal inertia tensor) then convert planar outputs into 3D sensor-equivalent signals and feed the geometry-induced increments back as external forces and moments at the center of gravity. This force and moment feedback is what changes the planar model's behavior, and the geometry transformation is what explains the yaw-rate correction, since yaw rate itself is not transformed.

What would settle it

Drive the same simulation on a road with laterally varying banking, such as a crowned or rutted road, and compare predicted vertical acceleration with measured values: because the ribbon cannot represent lateral banking variation, agreement should degrade exactly where the paper's synthetic and real tests do not probe. Alternatively, perturb the extracted spine geometry by a 0.5 m lateral offset or a 1-degree banking change and check whether the predicted corrections shift by a comparable amount; if they do, the reported error reductions depend on the accuracy of the geometry extraction from low-speed laps.

Watch

Extended reading notes

Core claim

The paper's central claim is that a planar vehicle dynamics model, treated as a black box, can be wrapped in a module that adapts it to 3D roads. The module does three things: it transforms the planar pose into a 3D pose on the road (geometry transformation), transforms the planar velocities, accelerations, and angular rates into their 3D equivalents including gravitational components (measurement transformation), and computes the extra forces and moments the road geometry induces, applied at the center of gravity. Because force and moment feedback changes what the planar model feels, the loop is closed. On the banked Las Vegas Motor Speedway, the approach reproduces measured accelerations and angular rates of a full-scale race car, where the planar baseline cannot; the largest error reductions are in vertical and lateral acceleration and yaw rate, and a closed-loop evaluation with a planar dual-track model and a motion controller remains stable. The authors do not claim to beat full 3D multi-body models, only to close most of the gap while keeping the planar model's speed and interpretability.

Load-bearing premise

The method stands on the road being representable as a ribbon—curvature, banking, and slope varying only with arc length—and on the vehicle staying close enough to the spine that the projection onto the flattened road plane is unambiguous; if the road banks differently across its width or the car drifts far laterally, the corrections inherit that error, and the paper provides no sensitivity analysis.

Editorial extensions

If this is right

  • Planar vehicle models can be used on real 3D roads without reformulating their equations, preserving their interpretability for control and state-estimation development.
  • The method runs at a mean 9.5 microseconds per step, making it usable for real-time simulation and sped-up training of reinforcement-learning agents.
  • Banked-road normal loads, which exceed the nominal load by over 66% on the Las Vegas track, are recovered well enough to alter tire-load-dependent dynamics in simulation.
  • Because the geometry transformation alone fixes most of the yaw-rate error, planar controllers evaluated on bird's-eye-view maps will see the road-plane behavior they would encounter in reality.
  • The open-source implementation lets other planar models be coupled as long as they accept external forces and moments as inputs.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The reported error reductions come from one track and one racing line, and the ribbon input is extracted from the same vehicle's low-speed laps, so geometry-extraction error is a plausible hidden source of bias in the validation numbers.
  • The same wrapping strategy should transfer to smooth, ribbon-representable roads such as highways and test tracks; roads with curbs, bumps, or laterally varying banking would need a superimposed local surface model, which the paper itself proposes as future work.
  • Because yaw rate is corrected by the geometry transformation rather than by force feedback, control algorithms tuned on flat-road simulations could be re-tuned on the road-plane reference line to match banked-oval behavior before track testing.
  • A natural extension is pairing the coupling with a tire model that explicitly uses the recovered normal load, since the 66% load increase would change friction limits and vehicle response in ways a point-mass validation cannot show.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper proposes a black-box coupling module that lets an existing planar vehicle dynamics model run on three-dimensional road geometry. The method consists of (i) a geometry transformation that projects the planar pose onto a ribbon-represented 3D road, (ii) transformations of velocity, angular velocity, acceleration, and angular acceleration into their 3D counterparts, and (iii) computation of geometry-induced forces and moments applied as external loads to the planar model. The authors validate on four synthetic tracks and on real high-speed data from a race car at Las Vegas Motor Speedway, reporting large reductions in mean absolute error versus a planar baseline, and they provide an open-source C++ implementation.

Significance. The core algebraic derivations in Sections III-D through III-F, including the cross-product expansion in Eq. (26) and the gravity projection in Eq. (27), are internally consistent, and the open-source implementation makes the method immediately usable and reproducible. If the real-data validation were quantitative for the closed-loop dynamic case, this would be a practically valuable contribution: it would let developers keep the interpretability and low cost of planar models while capturing banked-road and elevation effects that are otherwise missed. The synthetic-track experiments (Figs. 10-12) convincingly demonstrate that the transformation produces physically sensible signals in extreme geometry. However, the current real-data evidence is largely kinematic rather than dynamic, so the central claim about closing the gap between planar simulation and real 3D roads is only conditionally supported.

major comments (3)
  1. [Section IV-B, Table II] The quantitative validation is performed with a planar point-mass model that 'perfectly tracks the road plane reference line in an open-loop simulation' (Section IV, first paragraph). Because the trajectory is prescribed, the forces and moments from Section III-F (Eqs. (34) and (38)) do not feed back into the simulated state; the reported error reductions of about 95% for az, 78% for ay, and 82% for the yaw-rate bias in Table II therefore measure how well the kinematic measurement transformation plus extracted track geometry reconstruct the recorded IMU signals at constant 69.5 m/s, not how well the dynamic coupling preserves vehicle dynamics. The closed-loop dual-track evaluation in Figs. 15 and 16 is qualitative and explicitly not intended to replicate the real oscillations, and it is not scored in Table II. The abstract's statement that the gap between planar simulation and real-world 3D roads is closed thus needs either a quantitative closed-loop comparison or a more modestly scoped claim.
  2. [Table II, Section IV-B] Table II reports errors only over 10% to 45% of the lap, which the authors state is the portion where the 3D geometry effects are most pronounced. Restricting the comparison to this segment can inflate the apparent improvement and leaves the full-lap behavior unknown. Please provide full-lap metrics, or justify the segment choice by showing that the planar baseline is already accurate over the rest of the lap and that the conclusion is insensitive to the segment boundaries.
  3. [Sections III-B, III-D, and V] The validation inherits the ribbon track representation, with curvature, banking, and slope extracted from low-speed laps using an approach adapted from Perantoni and Limebeer [27]. The paper provides no sensitivity analysis of the predicted corrections to errors in this geometry extraction, nor to lateral deviations from the reference line, even though Section III-D itself notes that the road-plane projection must remain injective and Section V concedes that laterally varying and uneven roads break the ribbon assumption entirely. Since the corrections in Eqs. (29) and (34) are directly proportional to the geometric quantities, a small extraction error propagates linearly into the reported accelerations; at minimum, the authors should report the expected error level of the geometry extraction or run a perturbation study on the Las Vegas data.
minor comments (5)
  1. [Introduction] There is a typo in the fourth paragraph: 'Intertial Measurement Unit' should be 'Inertial Measurement Unit'.
  2. [Fig. 3] The description refers to 'modules drawn in blue', but the figure may not be distinguishable in grayscale; please use labels or a colorblind-safe palette in addition to color.
  3. [Table I] The vehicle parameters are said to be 'approximated from the recorded data'; please state how each parameter was estimated and, if possible, give an uncertainty range, since some predicted moments depend linearly on Ix, Iy, and Iz.
  4. [Section IV-B] The 2.0 Hz lowpass filter cutoff for the real-world IMU data is stated but not justified; a sentence explaining that the cutoff was chosen to remove sensor noise and vibration while preserving the track-induced dynamics would help reproducibility.
  5. [Section III-D] The small-angle approximation in Eqs. (9)-(10) is correct but would be clearer if it were stated as the leading-order Taylor expansion of (1-cos(alpha))/Omega_z,m and sin(alpha)/Omega_z,m, since the current text leaves the limit step implicit.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity; only minor non-load-bearing self-citations for the open-source dual-track model.

full rationale

The derivation chain in Sections III-D through III-F is self-contained: the geometry transformation preserves curvature and arc length and constructs the planar reference line from the measured curvature profile; the measurement transformation (Eqs. 20-30) is a forward kinematic mapping from track geometry and planar state to three-dimensional IMU quantities; and the force/moment feedback (Eqs. 32-38) is the difference between the spatial and planar rigid-body loads. The real-data validation uses track characteristics computed from independent low-speed laps (Fig. 14) to predict high-speed 69.5 m/s IMU data that were not used to fit the geometry, so the agreement in Table II is a cross-check of the forward model rather than a fit renamed as a prediction. The open-loop point-mass setup (Section IV-B) means the reported ME/MAE reductions do not exercise the dynamic feedback, and the closed-loop dual-track comparison is qualitative; this limits what the validation demonstrates, but it does not make any predicted quantity equal to its input by construction. The self-citations to the authors' openly available dual-track model ([4], [45]) are tool citations that are reproducible and not load-bearing to the central claim. Section V's stated limitations (rigid-body center-of-gravity collapsing, ribbon-only road shapes) are acknowledged scope restrictions, not hidden circular dependencies.

Assumptions & free parameters 5 free parameters · 7 assumptions · 0 invented entities

The central kinematic predictions (vertical and lateral acceleration corrections) are parameter-free functions of measured track geometry and vehicle state; the model inputs are the ribbon track representation, standard Newton-Euler rigid-body kinematics, and vehicle constants that the paper says were approximated from recorded data. No new physical entities are postulated. The virtual road plane is a mathematical embedding of the ribbon, not a claim about nature.

free parameters (5)
  • vehicle mass m = 800 kg
    Table I states it was approximated from the recorded data. Scales the applied force corrections Delta F = m (a - a_bar); does not affect the open-loop acceleration comparisons.
  • center-of-gravity height h = 0.3 m
    Table I, approximated from recorded data. Enters the small velocity and acceleration corrections (omega_y h, omega_dot_y h); subdominant for the Las Vegas signals.
  • moments of inertia Ix, Iy, Iz = 100 / 500 / 1000 kg m^2
    Table I, approximated from recorded data. Determine the moment corrections in Eq. (38) and hence the closed-loop roll, pitch, and yaw dynamics.
  • measurement filter cutoff = 2.0 Hz
    Fourth-order Butterworth applied only to the recorded data before computing ME and MAE in Table II; a hand-chosen analysis parameter that reduces the apparent oscillation-induced error.
  • planar reference line segment window = 100 points at 1 m spacing
    Section III-D: the road-plane reference line is computed in short segments and iteratively updated; segment length and update strategy are hand-chosen, with no convergence study.
assumptions (7)
  • domain assumption The 3D road is representable as a ribbon: a spine curve plus lateral offset, with curvature, banking, and slope that depend only on arc length.
    Section III-B adopts the Perantoni-Limebeer ribbon [27]. Section V concedes this cannot handle lateral banking changes, bumps, or curbs; those roads break the method's input stage.
  • domain assumption All road-induced loads can be collected at the center of gravity; suspension kinematics and gyroscopic moments are negligible.
    Section III-C justifies via sprung mass dominance [22], [45]; Section V lists it as the first principal limitation.
  • domain assumption The road-plane projection preserves curvature and arc length by construction and remains injective over the simulated trajectory.
    Equations (3)-(4) in Section III-D define the isometry; injectivity is handled only by truncating the reference line to 100-point segments, with no proof of seamlessness across segment joints.
  • standard math Standard Newton-Euler rigid-body kinematics with gravity projection are valid for the corrected 3D state.
    Section III-F, Eqs. (32) to (38); the cross-product expansion in Eq. (26) checks out term by term.
  • domain assumption Tire normal load is adequately represented by rigid-body vertical specific force, without a tire or suspension model in the quantitative validation.
    Table I and Section IV: the open-loop validation is a point-mass model mirroring mass and inertias; per-tire normal loads and load transfer are not modeled.
  • domain assumption The racing-line geometry extracted from low-speed laps of the same vehicle is accurate enough for predictions at 69.5 m/s.
    Fig. 14 uses the Perantoni-Limebeer identification approach [27]; no sensitivity analysis of the corrections to extraction error is provided.
  • ad hoc to paper Aerodynamic loads on the race car at about 250 km/h are negligible for the validation signals.
    Nowhere stated or modeled, yet the no-aero point-mass matches the IMU vertical acceleration to MAE 0.23 m/s^2; if aero were significant, the match would require explanation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Beyond the Plane: Coupling Planar Vehicle Dynamics with Three-Dimensional Road Geometry." pith.science (2026). https://pith.science/paper/QMAFTVS3

@misc{pith2026260809402,
  author       = {Pith},
  title        = {Pith review of: Beyond the Plane: Coupling Planar Vehicle Dynamics with Three-Dimensional Road Geometry},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QMAFTVS3}},
  note         = {Machine review of arXiv:2608.09402}
}
read the original abstract

Simulation is crucial for developing and testing autonomous driving systems. In particular, the development of localization and control algorithms relies on an accurate vehicle dynamics simulation. However, most vehicle dynamics models are two-dimensional while real-world roads are three-dimensional. For example, effects from the three-dimensional road geometry on the Las Vegas Motor Speedway can increase the normal forces on the tires by more than 66% compared to the nominal load at standstill. As a result, even highly detailed planar vehicle dynamics models struggle to accurately reproduce the real vehicle's behavior. While solutions for three-dimensional vehicle dynamics exist, they are rarely adopted, computationally expensive, and complex. To address this issue, we present a novel method to couple planar vehicle dynamics models with real-world three-dimensional road geometry. We transform the planar vehicle state from the vehicle model's two-dimensional plane to its corresponding representation in three-dimensional space. Additionally, we calculate road-geometry-induced forces and moments and apply them to the planar vehicle model. We validate our approach using high-speed data recorded with a full-scale race car on the banked Las Vegas Motor Speedway. Furthermore, on synthetic tracks, we show that our method yields accurate results even in edge cases. Together, our results demonstrate that the gap between planar simulation and real-world three-dimensional roads can be closed without abandoning simpler planar models. To simplify adoption of our method, we provide the implementation as open-source software on github.com/TUMFTM/3d-road-geometry-coupling.

Figures

Figures reproduced from arXiv: 2608.09402 by the authors.

Figure 1
Figure 1. Our AV21 race car used for data collection during an overtaking [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Relevant coordinate frames on the three-dimensional road surface. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Instead of modifying the model’s internals, we interact solely with the vehicle model through its input and output interfaces. We use the planar outputs from the vehicle model and transform them according to the car’s current state on the track to obtain the corresponding three-dimensional quantities. This is done in two distinct steps: First, we perform a geometry transformation to convert the vehicle’s planar pose… view at source ↗
Figures from the paper (10 more)
Figure 4
Figure 4. Figure 4: Conceptual illustration of the geometry transformation. The road [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]
Figure 6
Figure 6. Figure 6: Overview of the geometry transformation for constructing a three [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]
Figure 7
Figure 7. Figure 7: Overview of the synthetic validation tracks. The centerline, which [PITH_FULL_IMAGE:figures/full_fig_p008_7.png]
Figure 9
Figure 9. Figure 9: Resulting planar reference line I C for the various validation tracks. Since Tflat is purely two-dimensional, its reference line is identical to the bird’s-eye view of the centerline for all the synthetic validation tracks. The initial point for constructing the planar…
Figure 11
Figure 11. Figure 11: Three-dimensional vehicle states when driving around [PITH_FULL_IMAGE:figures/full_fig_p009_11.png]
Figure 13
Figure 13. Figure 13: Track layout of the Las Vegas Motor Speedway. The centerline is [PITH_FULL_IMAGE:figures/full_fig_p010_13.png]
Figure 14
Figure 14. Figure 14: Key characteristics of the Las Vegas Motor Speedway along the [PITH_FULL_IMAGE:figures/full_fig_p010_14.png]
Figure 15
Figure 15. Figure 15: Longitudinal, lateral, and vertical accelerations [PITH_FULL_IMAGE:figures/full_fig_p011_15.png]
Figure 17
Figure 17. Figure 17: Execution time of our method for 1 lap around the Las Vegas Motor [PITH_FULL_IMAGE:figures/full_fig_p012_17.png]
Figure 18
Figure 18. Figure 18: Progress over time for Tflat and Televated. The plot visualizes the temporal stretching between tracks that is hidden when signals are plotted over spatial progress. ACKNOWLEDGMENT Author contributions: Simon Sagmeister, as the first author, designed the structure of …

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

52 extracted references · 46 canonical work pages

  1. [27]

    Optimal Control of a Formula One Car on a Three-Dimensional Track—Part 1: Track Modeling and Identification,

    G. Perantoni and D. J. N. Limebeer, “Optimal Control of a Formula One Car on a Three-Dimensional Track—Part 1: Track Modeling and Identification,”Journal of Dynamic Systems, Measurement, and Control, vol. 137, no. 5, p. 051018, May 2015. [Online]. Available: https: //asmedigitalcollection.asme.org/dynamicsystems/article/doi/10.1115/ 1.4028253/370529/Optim...

    work page 2015

  2. [1]

    A comparative study on trajectory tracking control methods for automated vehicles,

    D. Zenget al., “A comparative study on trajectory tracking control methods for automated vehicles,”Scientific Reports, vol. 15, no. 1, p. 17073, May 2025. [Online]. Available: https://www.nature.com/articles/ s41598-025-01365-9

    work page 2025

  3. [2]

    Comparison of trajectory tracking controllers for autonomous vehicles,

    D. Calzolari, B. Schurmann, and M. Althoff, “Comparison of trajectory tracking controllers for autonomous vehicles,” in2017 IEEE 20th International Conference on Intelligent Transportation Systems (ITSC). Yokohama: IEEE, Oct. 2017, pp. 1–8. [Online]. Available: http://ieeexplore.ieee.org/document/8317800/

    work page arXiv 2017

  4. [3]

    Tube model predictive control for an autonomous race car,

    A. Wischnewski, M. Euler, S. G ¨um¨us, and B. Lohmann, “Tube model predictive control for an autonomous race car,”Vehicle System Dynamics, vol. 60, no. 9, pp. 3151–3173, Sep. 2022. [Online]. Available: https://www.tandfonline.com/doi/full/10.1080/00423114.2021.1943461

    arXiv 2022

  5. [4]

    Analyzing the Impact of Simulation Fidelity on the Evaluation of Autonomous Driving Motion Control,

    S. Sagmeister, P. Kounatidis, S. Goblirsch, and M. Lienkamp, “Analyzing the Impact of Simulation Fidelity on the Evaluation of Autonomous Driving Motion Control,” in2024 IEEE Intelligent Vehicles Symposium (IV). Jeju Island, Korea, Republic of: IEEE, Jun. 2024, pp. 230–237. [Online]. Available: https://ieeexplore.ieee.org/ document/10588858/

    work page arXiv 2024

  6. [5]

    Optimization-Based Autonomous Racing of 1:43 Scale RC Cars

    A. Liniger, A. Domahidi, and M. Morari, “Optimization-Based Autonomous Racing of 1:43 Scale RC Cars,”Optimal Control Applications and Methods, vol. 36, no. 5, pp. 628–647, Sep. 2015. [Online]. Available: http://arxiv.org/abs/1711.07300

    work page Pith review arXiv 2015

  7. [6]

    Stanley: The Robot that Won the DARPA Grand Challenge,

    S. Thrunet al., “Stanley: The Robot that Won the DARPA Grand Challenge,”Journal of Field Robotics, vol. 23, no. 9, pp. 661–692, Sep. 2006

    work page 2006

  8. [7]

    Indy Autonomous Challenge,

    P. Mitchell, “Indy Autonomous Challenge,” 2024. [Online]. Available: https://www.indyautonomouschallenge.com/

    work page 2024

Show all 52 references
  1. [8]

    Online Time-Optimal Trajectory Planning on Three-Dimensional Race Tracks,

    M. Rowold, L. ¨Ogretmen, U. Kasolowsky, and B. Lohmann, “Online Time-Optimal Trajectory Planning on Three-Dimensional Race Tracks,” in2023 IEEE Intelligent Vehicles Symposium (IV). Anchorage, AK, USA: IEEE, Jun. 2023, pp. 1–8. [Online]. Available: https://ieeexplore.ieee.org/d...

    2023

  2. [9]

    Optimal Control of a Formula One Car on a Three-Dimensional Track—Part 2: Optimal Control,

    D. J. N. Limebeer and G. Perantoni, “Optimal Control of a Formula One Car on a Three-Dimensional Track—Part 2: Optimal Control,”Journal of Dynamic Systems, Measurement, and Control, vol. 137, no. 5, p. 051019, May 2015. [Online]. Available: https: //asmedigitalcollection.asme....

    2015

  3. [10]

    Curved-ribbon-based track modelling for minimum lap-time optimisation,

    S. Lovato, M. Massaro, and D. J. N. Limebeer, “Curved-ribbon-based track modelling for minimum lap-time optimisation,”Meccanica, vol. 56, no. 8, pp. 2139–2152, Aug. 2021. [Online]. Available: https://link.springer.com/10.1007/s11012-021-01387-3

    2021 doi

  4. [11]

    Sampling- Based Motion Planning with Online Racing Line Generation for Autonomous Driving on Three-Dimensional Race Tracks,

    L. ¨Ogretmen, M. Rowold, A. Langmann, and B. Lohmann, “Sampling- Based Motion Planning with Online Racing Line Generation for Autonomous Driving on Three-Dimensional Race Tracks,” in2024 IEEE Intelligent Vehicles Symposium (IV). Jeju Island, Korea, Republic of: IEEE, Jun. 2024...

    2024

  5. [12]

    Modelling and control for high-sideslip vehicles on nonplanar roads,

    L. Ruifeng, R. Hongbin, and C. Chih-Keng, “Modelling and control for high-sideslip vehicles on nonplanar roads,”Vehicle System Dynamics, pp. 1–26, Apr. 2026. [Online]. Available: https: //www.tandfonline.com/doi/full/10.1080/00423114.2026.2660695

    2026

  6. [13]

    Non-euclidean vehicle motion models,

    T. D. Fork, “Non-euclidean vehicle motion models,” Ph.D. dissertation, University of California, Berkeley, 2024

    2024

  7. [14]

    Models and Predictive Control for Nonplanar Vehicle Navigation,

    T. Fork, H. E. Tseng, and F. Borrelli, “Models and Predictive Control for Nonplanar Vehicle Navigation,” in2021 IEEE International Intelligent Transportation Systems Conference (ITSC). Indianapolis, IN, USA: IEEE, Sep. 2021, pp. 749–754. [Online]. Available: https://ieeexplore...

    2021

  8. [15]

    Three-Dimensional Vehicle Dynamics State Estimation for High-Speed Race Cars under varying Signal Quality,

    S. Goblirsch, M. Weinmann, and J. Betz, “Three-Dimensional Vehicle Dynamics State Estimation for High-Speed Race Cars under varying Signal Quality,” in2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). Abu Dhabi, United Arab Emirates: IEEE, Oct. 2...

    2024

  9. [16]

    Vehicle dynamic estimation with road bank angle consideration for rollover detection: Theoretical and experimental studies,

    H. Dahmani, M. Chadli, A. Rabhi, and A. El Hajjaji, “Vehicle dynamic estimation with road bank angle consideration for rollover detection: Theoretical and experimental studies,”Vehicle System Dynamics, vol. 51, no. 12, pp. 1853–1871, Dec. 2013. [Online]. Available: http://www....

    2013

  10. [18]

    Motion planning and control for multi vehicle autonomous racing at high speeds,

    A. Rajiet al., “Motion planning and control for multi vehicle autonomous racing at high speeds,” inProc. IEEE Intell. Transp. Syst. Conf. (ITSC), Oct. 2022, pp. 2775–2782. [Online]. Available: https://doi.org/10.1109/ITSC55140.2022.9922239

    2022

  11. [19]

    A Tube- MPC Approach to Autonomous Multi-Vehicle Racing on High-Speed Ovals,

    A. Wischnewski, T. Herrmann, F. Werner, and B. Lohmann, “A Tube- MPC Approach to Autonomous Multi-Vehicle Racing on High-Speed Ovals,”IEEE Transactions on Intelligent Vehicles, pp. 1–1, 2022. [Online]. Available: https://ieeexplore.ieee.org/document/9762493/

    2022

  12. [20]

    Robust Nonlinear Trajectory Tracking Control for Autonomous Racing on Three-Dimensional Tracks,

    J. F. Bongard, G. Jank, S. Sagmeister, and B. Lohmann, “Robust Nonlinear Trajectory Tracking Control for Autonomous Racing on Three-Dimensional Tracks,” 2026. [Online]. Available: https: //arxiv.org/abs/2604.19452

    2026 arXiv

  13. [21]

    High-Speed, All-Terrain Autonomy: Ensuring Safety at the Limits of Mobility,

    J. R. Baxter, B. I. Epureanu, P. Jayakumar, and T. Ersal, “High-Speed, All-Terrain Autonomy: Ensuring Safety at the Limits of Mobility,”

  14. [22]

    W. F. Milliken and D. L. Milliken,Race Car Vehicle Dynamics. Warrendale, PA, U.S.A: SAE International, 1995

    1995

  15. [23]

    Allen, H

    R. Allen, H. T. Szostak, D. H. Klyde, T. J. Rosenthal, and K. J. Owens,Vehicle Dynamic Stability and Rollover. Washington, D.C.: National Highway Traffic Safety Administration,

  16. [24]

    T. D. Gillespie,Fundamentals of Vehicle Dynamics. Warrendale, PA: SAE International, Feb. 1992. [Online]. Available: https: //saemobilus.sae.org/content/R-114/

    1992

  17. [25]

    Models for ground vehicle control on nonplanar surfaces,

    T. Fork, H. E. Tseng, and F. Borrelli, “Models for ground vehicle control on nonplanar surfaces,”Vehicle System Dynamics, vol. 62, no. 4, pp. 976–1000, Apr. 2024. [Online]. Available: https://www.tandfonline.com/doi/full/10.1080/00423114.2023.2205596

    2024

  18. [26]

    A General 3D Road Model for Motorcycle Racing,

    T. Fork and F. Borrelli, “A General 3D Road Model for Motorcycle Racing,” in16th International Symposium on Advanced Vehicle Control, G. Mastinu, F. Braghin, F. Cheli, M. Corno, and S. M. Savaresi, Eds. Cham: Springer Nature Switzerland, 2024, pp. 357–363. [Online]. Available:...

    2024 doi

  19. [28]

    Realistic 3D Simulators for Automotive: A Review of Main Applications and Features,

    I. Silva, H. Silva, F. Botelho, and C. Pend ˜ao, “Realistic 3D Simulators for Automotive: A Review of Main Applications and Features,” Sensors, vol. 24, no. 18, p. 5880, Sep. 2024. [Online]. Available: https://www.mdpi.com/1424-8220/24/18/5880

    2024

  20. [29]

    A Survey on Simulators for Testing Self-Driving Cars,

    P. Kaur, S. Taghavi, Z. Tian, and W. Shi, “A Survey on Simulators for Testing Self-Driving Cars,” in2021 Fourth International Conference on Connected and Autonomous Driving (MetroCAD). Detroit, MI, USA: IEEE, Apr. 2021, pp. 62–70. [Online]. Available: https://ieeexplore.ieee.o...

    2021

  21. [30]

    Terrain and Road Surfaces

    CarSim, “Terrain and Road Surfaces.” [Online]. Available: https: //www.carsim.com/products/supporting/terrain/index.php

  22. [31]

    A Generic Multibody Vehicle Model for Simulating Handling and Braking,

    M. W. Sayers and D. Han, “A Generic Multibody Vehicle Model for Simulating Handling and Braking,”Vehicle System Dynamics, vol. 25, no. sup1, pp. 599–613, Jan. 1996. [Online]. Available: http://www.tandfonline.com/doi/abs/10.1080/00423119608969223

    1996 doi

  23. [32]

    Road Characterization for the Simulation of Automotive Vehicle Dynamics,

    M. W. Sayers, “Road Characterization for the Simulation of Automotive Vehicle Dynamics,”SAE International Journal of Materials and Manufacturing, vol. 04, no. 1, pp. 251–262, Apr. 2011. [Online]. Available: https://saemobilus.sae.org/articles/road-characterization- simulation-...

    2011

  24. [33]

    The Vehicle Dynamics Library: New Concepts and New Fields of Application,

    J. Andreasson, “The Vehicle Dynamics Library: New Concepts and New Fields of Application,” inThe 8th International Modelica Conference, Technical Univeristy, Dresden, Germany, Jun. 2011, pp. 414–420. [Online]. Available: https://ep.liu.se/en/conference-article.aspx?series= ecp...

    2011

  25. [34]

    Hybrid Vehicle Model Development using ASM-AMESim-Simscape Co-Simulation for Real-Time HIL Applications,

    K. Patil, S. K. Molla, and T. Schulze, “Hybrid Vehicle Model Development using ASM-AMESim-Simscape Co-Simulation for Real-Time HIL Applications,” inSAE 2012 World Congress & Exhibition, Detroit, Michigan, United States, Apr. 2012, pp. 2012–01–0932. [Online]. Available: https:/...

    2012

  26. [35]

    Choose Your Simulator Wisely: A Review on Open-Source Simulators for Autonomous Driving,

    Y . Liet al., “Choose Your Simulator Wisely: A Review on Open-Source Simulators for Autonomous Driving,”IEEE Transactions on Intelligent SAGMEISTERet al.: BEYOND THE PLANE: COUPLING PLANAR VEHICLE DYNAMICS WITH THREE-DIMENSIONAL ROAD GEOMETRY 15 Vehicles, vol. 9, no. 5, pp. 48...

    2024

  27. [36]

    A Simulation Benchmark for Autonomous Racing with Large-Scale Human Data,

    M. Bertognaet al., “A Simulation Benchmark for Autonomous Racing with Large-Scale Human Data,” inAdvances in Neural Information Processing Systems 37. Vancouver, BC, Canada: Neural Information Processing Systems Foundation, Inc. (NeurIPS), 2024, pp. 102 078–102 100. [Online]. ...

    2024

  28. [37]

    Outracing champion Gran Turismo drivers with deep reinforcement learning,

    P. R. Wurmanet al., “Outracing champion Gran Turismo drivers with deep reinforcement learning,”Nature, vol. 602, no. 7896, pp. 223–228, Feb. 2022. [Online]. Available: https://www.nature.com/ articles/s41586-021-04357-7

    2022

  29. [38]

    Multi-body Vehicle Dynamics Simulation Based on Measured 3-D Terrain Data,

    T. Varunjikar, P. K. Vemulapalli, and S. N. Brennan, “Multi-body Vehicle Dynamics Simulation Based on Measured 3-D Terrain Data,” in3rd International Conference on Road Safety and Simulation, Inidianapolis, 2011. [Online]. Available: http://onlinepubs.trb.org/ onlinepubs/confe...

    2011

  30. [39]

    An Effective Algorithm of Uneven Road Surface Modeling and Calculating Reaction Forces for a Vehicle Dynamics Simulation,

    S. Tengler and K. Warwas, “An Effective Algorithm of Uneven Road Surface Modeling and Calculating Reaction Forces for a Vehicle Dynamics Simulation,”Coatings, vol. 11, no. 5, p. 535, Apr. 2021. [Online]. Available: https://www.mdpi.com/2079-6412/11/5/535

    2021

  31. [40]

    CARLA: An open urban driving simulator,

    A. Dosovitskiy, G. Ros, F. Codevilla, A. Lopez, and V . Koltun, “CARLA: An open urban driving simulator,” inProceedings of the 1st Annual Conference on Robot Learning, 2017, pp. 1–16

    2017

  32. [41]

    R-CARLA: High-Fidelity Sensor Simulations with Interchangeable Dynamics for Autonomous Racing,

    M. Brunner, E. Ghignone, N. Baumann, and M. Magno, “R-CARLA: High-Fidelity Sensor Simulations with Interchangeable Dynamics for Autonomous Racing,” in2025 IEEE Intelligent Vehicles Symposium (IV). Cluj-Napoca, Romania: IEEE, Jun. 2025, pp. 2558–2564. [Online]. Available: https...

    2025

  33. [42]

    Chrono: An Open Source Multi-physics Dynamics Engine,

    A. Tasoraet al., “Chrono: An Open Source Multi-physics Dynamics Engine,” inHigh Performance Computing in Science and Engineering, T. Kozubek, R. Blaheta, J. ˇS´ıstek, M. Rozlo ˇzn´ık, and M. ˇCerm´ak, Eds. Cham: Springer International Publishing, 2016, vol. 9611, pp. 19–49. [O...

    2016 doi

  34. [43]

    Chrono::Vehicle: Template-based ground vehicle modelling and simulation,

    R. Serban, M. Taylor, D. Negrut, and A. Tasora, “Chrono::Vehicle: Template-based ground vehicle modelling and simulation,”International Journal of Vehicle Performance, vol. 5, no. 1, p. 18, 2019. [Online]. Available: http://www.inderscience.com/link.php?id=97096

    2019

  35. [44]

    On Determinism of Game Engines Used for Simulation-Based Autonomous Vehicle Verification,

    G. Chanceet al., “On Determinism of Game Engines Used for Simulation-Based Autonomous Vehicle Verification,”IEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 11, pp. 20 538– 20 552, Nov. 2022. [Online]. Available: https://ieeexplore.ieee.org/ document/9793395/

    2022

  36. [45]

    Open Car Dynamics,

    S. Sagmeister, S. Hoffmann, and P. Kounatidis, “Open Car Dynamics,”

  37. [46]

    Robust and Efficient Curvilinear Coordinate Transformation with Guaranteed Map Coverage for Motion Planning,

    G. W ¨ursching and M. Althoff, “Robust and Efficient Curvilinear Coordinate Transformation with Guaranteed Map Coverage for Motion Planning,” in2024 IEEE Intelligent Vehicles Symposium (IV). Jeju Island, Korea, Republic of: IEEE, Jun. 2024, pp. 2694–2701. [Online]. Available: ...

    2024

  38. [47]

    Lanelets: Efficient map representation for autonomous driving,

    P. Bender, J. Ziegler, and C. Stiller, “Lanelets: Efficient map representation for autonomous driving,” in2014 IEEE Intelligent Vehicles Symposium Proceedings. MI, USA: IEEE, Jun. 2014, pp. 420–

    2014

  39. [48]

    CommonRoad Drivability Checker: Simplifying the Development and Validation of Motion Planning Algorithms,

    C. Pek, V . Rusinov, S. Manzinger, M. C. Uste, and M. Althoff, “CommonRoad Drivability Checker: Simplifying the Development and Validation of Motion Planning Algorithms,” in2020 IEEE Intelligent Vehicles Symposium (IV). Las Vegas, NV , USA: IEEE, Oct. 2020, pp. 1013–1020. [Onl...

    2020

  40. [49]

    On the Theory of Filter Amplifiers,

    S. Butterworth, “On the Theory of Filter Amplifiers,”Experimental Wireless & the Wireless Engineer, vol. 7, pp. 536–541, Oct. 1930

    1930

  41. [425]

    Available: http://ieeexplore.ieee.org/document/6856487/

    [Online]. Available: http://ieeexplore.ieee.org/document/6856487/

  42. [1992]

    Available: https://babel.hathitrust.org/cgi/pt?id= mdp.39015075298698&view=1up&seq=119&skin=2021

    [Online]. Available: https://babel.hathitrust.org/cgi/pt?id= mdp.39015075298698&view=1up&seq=119&skin=2021

    2021

  43. [2024]

    Available: https://github.com/TUMFTM/Open-Car- Dynamics

    [Online]. Available: https://github.com/TUMFTM/Open-Car- Dynamics

  44. [2026]

    Available: https://arxiv.org/abs/2603.20525

    [Online]. Available: https://arxiv.org/abs/2603.20525

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.