pith. sign in

arxiv: 2511.02006 · v1 · submitted 2025-11-03 · 📡 eess.SP

A Comparison of Road Grade Preview Signals from Lidar and Maps

Logan Schexnaydre , Aman Poovalappil , Darrell Robinette , Jeremy Bos This is my paper

Pith reviewed 2026-05-18 00:55 UTC · model grok-4.3

classification 📡 eess.SP
keywords road grade estimationlidarautonomous vehiclespreview controlmap comparisonKalman filterroad surface measurement
0
0 comments X

The pith

Lidar sensors estimate upcoming road grade with unbiased 0.6-degree error at 53 meters ahead, matching map precision for vehicle control.

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

The paper shows that road grade preview, which matters for energy efficiency, safety, and comfort in automated driving, does not have to wait until the vehicle drives over the segment or depend entirely on pre-made maps. Instead, onboard lidar can accumulate point returns while moving and use Kalman filtering to estimate the height difference across the wheelbase at upcoming waypoints along the path. A field experiment compares these estimates directly to a GNSS/INS reference map and finds an unbiased error whose standard deviation is 0.6 degrees at an average range of 52.7 meters. This result indicates that lidar supplies a practical real-time alternative when maps are unavailable or inaccurate, giving control systems an independent preview signal.

Core claim

Through field testing on real roads, the lidar-based estimator that accumulates point-cloud returns and applies Kalman filtering to reduce odometry uncertainty produces road-grade values whose difference from a high-accuracy GNSS/INS map has zero mean bias and a standard deviation of 0.6 degrees, observed at an average preview distance of 52.7 meters. The paper therefore concludes that automotive lidar constitutes a valid source of grade preview for automated-vehicle control tasks.

What carries the argument

Lidar grade estimator that builds height differences between front and rear wheelbase positions from accumulated point returns at each waypoint, using Kalman filtering to limit the impact of motion and odometry uncertainty.

If this is right

  • Control systems can begin compensating for grade before the vehicle reaches the slope instead of reacting after the fact.
  • Automated vehicles obtain an extra, independent preview channel that continues to function when map data is missing or stale.
  • Grade-aware energy management and stability algorithms become feasible in unmapped or rapidly changing environments.
  • Sensor redundancy increases because the same lidar unit already used for object detection can also supply road geometry.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same accumulation-and-filter approach could be fused with camera or radar returns to maintain performance when lidar returns are sparse.
  • Preview range and accuracy might improve further by incorporating vehicle speed and suspension data into the filter model.
  • Widespread adoption would reduce the need for continuous map updates in regions with frequent construction or seasonal changes.

Load-bearing premise

The GNSS/INS map is accepted as perfectly accurate ground truth and the chosen test roads and conditions are taken to represent situations where lidar returns remain reliable.

What would settle it

An independent high-precision survey of road grade along the same paths using total-station or similar equipment; if the standard deviation of lidar-minus-survey errors exceeds roughly 1 degree or shows clear bias, the claim of comparable precision would be refuted.

Figures

Figures reproduced from arXiv: 2511.02006 by Aman Poovalappil, Darrell Robinette, Jeremy Bos, Logan Schexnaydre.

Figure 1
Figure 1. Figure 1: At each waypoint, the difference in height between the front and [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The process used to create filtered grade estimates. The magenta [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: A bird’s-eye view of the vehicle front and rear contact patch locations [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: The path recorded by the NovAtel PwrPak7 is a single loop where [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: The elevation and grade angle recorded by the NovAtel PwrPak7 at [PITH_FULL_IMAGE:figures/full_fig_p004_5.png] view at source ↗
Figure 7
Figure 7. Figure 7: A comparison of the grade estimated by the Kalman-filtered lidar [PITH_FULL_IMAGE:figures/full_fig_p005_7.png] view at source ↗
Figure 6
Figure 6. Figure 6: The error of the unfiltered lidar-based grade estimator as the number [PITH_FULL_IMAGE:figures/full_fig_p005_6.png] view at source ↗
Figure 8
Figure 8. Figure 8: The error in the Kalman-filtered lidar-based estimator when being [PITH_FULL_IMAGE:figures/full_fig_p005_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: The residual of the Kalman filter applied to the lidar-based grade [PITH_FULL_IMAGE:figures/full_fig_p006_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: The absolute Kalman-filtered grade estimation error with respect to [PITH_FULL_IMAGE:figures/full_fig_p006_10.png] view at source ↗
read the original abstract

Road grade can impact the energy efficiency, safety, and comfort associated with automated vehicle control systems. Currently, control systems that attempt to compensate for road grade are designed with one of two assumptions. Either the grade is only known once the vehicle is driving over the road segment through proprioception, or complete knowledge of the oncoming road grade is known from a pre-made map. Both assumptions limit the performance of a control system, as not having a preview signal prevents proactive grade compensation, whereas relying only on map data potentially subjects the control system to missing or outdated information. These limits can be avoided by measuring the oncoming grade in real-time using on-board lidar sensors. In this work, we use point returns accumulated during travel to estimate the grade at each waypoint along a path. The estimated grade is defined as the difference in height between the front and rear wheelbase at a given waypoint. Kalman filtering techniques are used to mitigate the effects of odometry and motion uncertainty on the grade estimates. This estimator's performance is compared to the measurements of a map created with a GNSS/INS system via a field experiment. When compared to the map-based system, the lidar-based estimator produces an unbiased error with a standard deviation of 0.6 degrees at an average range of 52.7 meters. By having similar precision to map-based systems, automotive lidar-based grade estimation systems are shown to be a valid approach for measuring road grade when a map is unavailable or inaccurate. In using lidar as an input signal for grade-based control system tasks, autonomous vehicles achieve higher redundancy and independence in contrast to existing methods.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 2 minor

Summary. The paper claims to demonstrate a lidar-based road grade preview estimator that accumulates point returns, applies Kalman filtering to mitigate odometry uncertainty, and produces grade estimates defined as height differences between front and rear wheelbase at waypoints. In a field experiment, this estimator is compared to a GNSS/INS-derived map and reported to yield an unbiased error with 0.6° standard deviation at an average range of 52.7 m, establishing lidar as a valid real-time alternative when maps are unavailable or inaccurate.

Significance. If the central empirical comparison holds after addressing validation gaps, the work supplies a concrete, sensor-based preview signal that increases redundancy for grade-compensating controllers in automated vehicles, directly addressing limitations of both proprioceptive and map-only approaches.

major comments (1)
  1. [Methods] Methods section (description of field experiment and comparison): the lidar estimates are subtracted directly from the GNSS/INS map at each waypoint to obtain the reported 0.6° std-dev error, yet no independent error budget, repeated-run consistency check, or cross-validation is provided for the map itself. Residual GNSS multipath, INS drift, or interpolation artifacts in the map could therefore be misattributed to the lidar estimator, rendering the precision claim and the conclusion that lidar is a valid substitute load-bearing on an untested assumption.
minor comments (2)
  1. [Abstract] Abstract: the phrase 'average range of 52.7 meters' should be defined explicitly (e.g., mean preview distance along the path or mean sensor-to-waypoint distance) to avoid ambiguity in interpreting the operating regime.
  2. [Methods] The Kalman-filter formulation for handling motion uncertainty is described at a high level; adding the state vector, process model, or covariance tuning details would improve reproducibility without altering the central claim.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful reading and constructive feedback on our manuscript. The point raised about the reference map validation is well taken, and we address it directly below with a commitment to revise the methods section.

read point-by-point responses

  1. Referee: [Methods] Methods section (description of field experiment and comparison): the lidar estimates are subtracted directly from the GNSS/INS map at each waypoint to obtain the reported 0.6° std-dev error, yet no independent error budget, repeated-run consistency check, or cross-validation is provided for the map itself. Residual GNSS multipath, INS drift, or interpolation artifacts in the map could therefore be misattributed to the lidar estimator, rendering the precision claim and the conclusion that lidar is a valid substitute load-bearing on an untested assumption.

    Authors: We agree that the comparison relies on the GNSS/INS map serving as a reference and that an explicit error budget for this reference was not provided in the original manuscript. The field experiment used a commercial-grade GNSS/INS unit operated in open-sky conditions to reduce multipath, and the reported 0.6° standard deviation reflects the observed difference between the two systems. To address the concern, the revised manuscript will include manufacturer-specified accuracy figures for the GNSS/INS (typically <0.1° attitude and cm-level position under good conditions) together with a brief propagation analysis showing how these translate to grade uncertainty at the reported ranges. We will also note that repeated-run consistency checks were not performed in this study due to the single-pass nature of the experiment but will add a limitations paragraph acknowledging that residual reference errors could contribute to the observed difference. This revision clarifies the assumption without altering the core empirical result or the conclusion that lidar provides a viable real-time alternative. revision: yes

Circularity Check

0 steps flagged

Empirical comparison to external GNSS/INS reference exhibits no circularity

full rationale

The paper reports results from a field experiment in which lidar point clouds are accumulated, Kalman-filtered to produce grade estimates defined as front-to-rear height differences at waypoints, and then subtracted from a separately constructed GNSS/INS map. The headline statistics (unbiased error, 0.6° standard deviation at 52.7 m average range) are direct numerical outcomes of this subtraction against an external reference; they are not obtained by fitting a parameter to the target quantity and re-using it as a prediction, nor by any self-citation chain, self-definition, or ansatz smuggling. Because the central claim rests on an independent external benchmark rather than reducing to the paper’s own inputs by construction, the derivation chain is self-contained.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The paper rests on standard assumptions about sensor noise and vehicle motion rather than introducing new free parameters or postulated entities; the central claim is supported by field data rather than derivation.

axioms (1)
  • domain assumption Kalman filter process and measurement noise models adequately capture odometry and motion uncertainty for grade estimation.
    Invoked to mitigate effects of uncertainty on the grade estimates.

pith-pipeline@v0.9.0 · 5826 in / 1307 out tokens · 38465 ms · 2026-05-18T00:55:46.610088+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

24 extracted references · 24 canonical work pages

  1. [1]

    Safe Adaptive Cruise Control with Road Grade Preview and \mathbfV2\mathbfV Communication,

    R. Firoozi, S. Nazari, J. Guanetti, R. O’Gorman, and F. Borrelli, “Safe Adaptive Cruise Control with Road Grade Preview and \mathbfV2\mathbfV Communication,” in2019 American Control Conference (ACC), pp. 4448–4453, July 2019. ISSN: 2378-5861

    work page 2019

  2. [2]

    Look-ahead control for heavy trucks to minimize trip time and fuel consumption,

    E. Hellstr ¨om, M. Ivarsson, J. ˚Aslund, and L. Nielsen, “Look-ahead control for heavy trucks to minimize trip time and fuel consumption,”Control Engineering Practice, vol. 17, pp. 245–254, Feb. 2009

    work page 2009

  3. [3]

    Improving cruise control efficiency through speed flexibility & on-board data,

    A. Grewal and M. Zebiak, “Improving cruise control efficiency through speed flexibility & on-board data,” tech. rep., SAE Technical Paper, 2023

    work page 2023

  4. [4]

    Role of Terrain Preview in Energy Management of Hybrid Electric Vehicles,

    C. Zhang, A. Vahidi, P. Pisu, X. Li, and K. Tennant, “Role of Terrain Preview in Energy Management of Hybrid Electric Vehicles,”IEEE Transactions on V ehicular Technology, vol. 59, pp. 1139–1147, Mar. 2010. MAG ID: 2142809728

    work page 2010

  5. [5]

    Road Profile Estimation Using a 3D Sensor and Intelligent Vehicle,

    T. Ni, W. Li, D. Zhao, and Z. Kong, “Road Profile Estimation Using a 3D Sensor and Intelligent Vehicle,”Sensors, vol. 20, p. 3676, Jan. 2020

    work page 2020

  6. [6]

    Road Grade and Vehicle Parameter Estimation for Longitudinal Control Using GPS,

    H. S. Bae, J. Ryu, and J. C. Gerdes, “Road Grade and Vehicle Parameter Estimation for Longitudinal Control Using GPS,” in Proceedings on the IEEE Conference on Intelligent Transportation Systems, pp. 25–29, 2001

    work page 2001

  7. [7]

    Real-Time Dynamic Mass Learning Algorithm for Connected and Automated Light-Duty Vehicles,

    A. Poovalappil, A. Robare, P. Apostol, M. Bahramgiri, B. Chen, J. Naber, and D. Robinette, “Real-Time Dynamic Mass Learning Algorithm for Connected and Automated Light-Duty Vehicles,” inWCX SAE World Congress Experience, SAE International, Apr. 2025

    work page 2025

  8. [8]

    A Novel Kalman Filter Based Road Grade Estimation Method,

    C. Gokcek, “A Novel Kalman Filter Based Road Grade Estimation Method,” SAE International, Apr. 2020. ISSN: 0148-7191, 2688-3627

    work page 2020

  9. [9]

    Road Grade Estimation With Vehicle-Based Inertial Measurement Unit and Orientation Filter,

    J. Jauch, J. Masino, T. Staiger, and F. Gauterin, “Road Grade Estimation With Vehicle-Based Inertial Measurement Unit and Orientation Filter,”IEEE Sensors Journal, vol. 18, pp. 781–789, Jan. 2018

    work page 2018

  10. [10]

    Road grade estimation for look-ahead vehicle control using multiple measurement runs,

    P. Sahlholm and K. Henrik Johansson, “Road grade estimation for look-ahead vehicle control using multiple measurement runs,”Control Engineering Practice, vol. 18, pp. 1328–1341, Nov. 2010

    work page 2010

  11. [11]

    Cooperative Estimation of Road Grade Based on Multidata Fusion for Vehicle Platoon with Optimal Energy Consumption,

    F. Ma, Y . Yang, J. Wang, Y . Zhao, Y . Shen, G. Wu, B. A. Guvenc, and L. Guvenc, “Cooperative Estimation of Road Grade Based on Multidata Fusion for Vehicle Platoon with Optimal Energy Consumption,” SAE International, Apr. 2020. ISSN: 0148-7191, 2688-3627

    work page 2020

  12. [12]

    Road Grade Estimation for On-Road Vehicle Emissions Modeling Using Light Detection and Ranging Data,

    K. Zhang and H. C. Frey, “Road Grade Estimation for On-Road Vehicle Emissions Modeling Using Light Detection and Ranging Data,”Journal of the Air & Waste Management Association, vol. 56, pp. 777–788, June 2006

    work page 2006

  13. [13]

    A Fully Automated Approach to Extract and Assess Road Cross Sections From Mobile LiDAR Data,

    S. A. Gargoum, K. El-Basyouny, K. Froese, and A. Gadowski, “A Fully Automated Approach to Extract and Assess Road Cross Sections From Mobile LiDAR Data,”IEEE Transactions on Intelligent Transportation Systems, vol. 19, pp. 3507–3516, Nov. 2018

    work page 2018

  14. [14]

    Mapping for Autonomous Driving: Opportunities and Challenges,

    K. Wong, Y . Gu, and S. Kamijo, “Mapping for Autonomous Driving: Opportunities and Challenges,”IEEE Intelligent Transportation Systems Magazine, vol. 13, no. 1, pp. 91–106, 2021

    work page 2021

  15. [15]

    Extraction of preview elevation of road based on 3D sensor,

    D. Zhao, L. Wang, Y . Li, and M. Du, “Extraction of preview elevation of road based on 3D sensor,”Measurement, vol. 127, pp. 104–114, Oct. 2018

    work page 2018

  16. [16]

    Extraction of Preview Elevation Information Based on Terrain Mapping and Trajectory Prediction in Real-Time,

    L. Wang, D. Zhao, T. Ni, and S. Liu, “Extraction of Preview Elevation Information Based on Terrain Mapping and Trajectory Prediction in Real-Time,”IEEE Access, vol. 8, pp. 76618–76631, 2020

    work page 2020

  17. [17]

    PointPillars: Fast Encoders for Object Detection From Point Clouds,

    A. H. Lang, S. V ora, H. Caesar, L. Zhou, J. Yang, and O. Beijbom, “PointPillars: Fast Encoders for Object Detection From Point Clouds,” in2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), (Long Beach, CA, USA), pp. 12689–12697, IEEE, June 2019

    work page 2019

  18. [18]

    GndNet: Fast Ground Plane Estimation and Point Cloud Segmentation for Autonomous Vehicles,

    A. Paigwar, O. Erkent, D. Sierra-Gonzalez, and C. Laugier, “GndNet: Fast Ground Plane Estimation and Point Cloud Segmentation for Autonomous Vehicles,” in2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 2150–2156, Oct. 2020. ISSN: 2153-0866

    work page 2020

  19. [19]

    Progressive LiDAR adaptation for road detection,

    Z. Chen, J. Zhang, and D. Tao, “Progressive LiDAR adaptation for road detection,”IEEE/CAA Journal of Automatica Sinica, vol. 6, pp. 693–702, May 2019

    work page 2019

  20. [20]

    Patchwork: Concentric Zone-Based Region-Wise Ground Segmentation With Ground Likelihood Estimation Using a 3D LiDAR Sensor,

    H. Lim, M. Oh, and H. Myung, “Patchwork: Concentric Zone-Based Region-Wise Ground Segmentation With Ground Likelihood Estimation Using a 3D LiDAR Sensor,”IEEE Robotics and Automation Letters, vol. 6, pp. 6458–6465, Oct. 2021

    work page 2021

  21. [21]

    LiDAR-Based Predictive Cruise Control,

    H. Alzu’bi, A. T. Jarbo, Q. Alrousan, and T. Tasky, “LiDAR-Based Predictive Cruise Control,” SAE International, Apr. 2020. ISSN: 0148-7191, 2688-3627

    work page 2020

  22. [22]

    A quantitative evaluation of surface normal estimation in point clouds,

    K. Jordan and P. Mordohai, “A quantitative evaluation of surface normal estimation in point clouds,” in2014 IEEE/RSJ International Conference on Intelligent Robots and Systems, pp. 4220–4226, Sept. 2014. ISSN: 2153-0866

    work page 2014

  23. [23]

    Estimating the Odometry Error of a Mobile Robot during Navigation,

    A. Martinelli and R. Siegwart, “Estimating the Odometry Error of a Mobile Robot during Navigation,” inProceedings of the 1st European Conference on Mobile Robots, (Radziejowice, Poland), p. 6 p., ZTUREK Research-Scientific Inst., Sept. 2003

    work page 2003

  24. [24]

    P. S. Maybeck,Stochastic Models, Estimation, and Control, vol. 1 ofMathematics in Science and Engineering. Academic Press, Inc., 1979. Contact Information Logan Schexnaydre lpschexn@mtu.edu Acknowledgments This work was supported by the ARPA-E NEXTCAR Program (Award number: DE-AR0000788). Definitions, Acronyms, Abbreviations 7 ADASAdvanced Driver Assistan...

    work page 1979