Pith. sign in

REVIEW 4 major objections 7 minor 19 references

Automotive Speed Estimation: Sensor Types and Error Characteristics from OBD-II to ADAS

T0 review · 4 major / 7 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read This paper claims that a vehicle's speed sensor type can be identified from OBD-II logs by whether speed ever reports below 3 km/h, and that the resulting speed error variance follows a two-term exponential model fitted from three urban…

desk verdict A useful, honest, single-vehicle study that proposes a plausible sensor-type heuristic and a speed-dependent error variance fit for OBD-II speed; the central generalizability claim is not yet established but the paper deserves serious refereeing. read the letter →

arxiv 2501.00242 v3 pith:YNSQSER4 submitted 2024-12-31 eess.SP cs.RO

Hany Ragab (1) , Sidney Givigi (2) , Aboelmagd Noureldin (1 , 2) ((1) Dept. of Electrical , Computer Engineering at Queens University , the NavINST Lab at the Royal Military College of Canada , (2) School of Computing at Queens University) This is my paper · ORCID
classification eess.SPcs.RO
keywords OBD-IIwheelspeedsensorpassivevsactiveWSSerrormodelingvehicleestimationstereovisualodometryfusionGNSS-deniednavigation
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

Modern navigation systems often draw vehicle speed from the OBD-II port, but the sensor behind that number varies by car, which changes the error statistics. This paper proposes a simple tell: if OBD-II speed readings never appear below 3 km/h, the vehicle uses a passive wheel speed sensor; active sensors report down to 0.1 km/h. Based on three real urban drives in a Toyota Sienna, the authors fit the passive sensor's error variance as $\sigma^2(v_f) = 0.3095 e^{-0.1241 v_f} + 0.1477 e^{-0.0009149 v_f}$, so errors are largest at low speed. They also show stereo visual odometry with semantic outlier rejection beats the OBD-II passive sensor below 4 km/h, while the OBD-II speed is more reliable in scenes full of static objects, arguing for a fusion engine. If right, an integrator can classify the sensor and schedule fusion weights from logs alone, improving dead-reckoning in GNSS-denied conditions.

What carries the argument

The load-bearing object is the OBD-II speed PID and the reporting threshold. Equation (1) defines the OBD-II speed as the integer-truncated vehicle speed estimate, and Eq. (4) shows the ABS-based estimate as the average of four individually computed wheel speeds. The classification rule uses the known passive-WSS suppression threshold of 3 km/h versus active WSS sensitivity to 0.1 km/h. The error model is the two-term exponential fit of Eq. (5), $\sigma^2_{PWSS}(v_f) = 0.3095 e^{-0.1241 v_f} + 0.1477 e^{-0.0009149 v_f}$, with coefficients determined by an optimization with 95% confidence bounds. That function supplies the speed-dependent covariance needed for sensor fusion.

What would settle it

Take a vehicle whose wheel speed sensor is known from the manufacturer to be active (for example, specified to detect 0.1 km/h) and log OBD-II PID 0x0D through stop-and-go driving; if the log contains no speed values below 3 km/h, the paper's classification rule would mislabel it as passive, falsifying the identification heuristic. A second check: compare the fitted variance model against a calibrated reference in a GNSS-challenged urban canyon; if the NovAtel reference itself carries bias there, the variance curve would be displaced.

Watch

Extended reading notes

Core claim

The central claim is that the sensor type and error behavior of an OBD-II speed signal can be recovered without any OEM cooperation. The diagnostic protocol reports only an integer vehicle speed (PID 0x0D, the truncation of the underlying estimate) and does not say whether the estimate came from a transmission-mounted vehicle speed sensor or from averaged ABS wheel speed sensors, nor whether those wheel sensors are passive or active. The paper's identification rule is that passive wheel speed sensors do not report below 3 km/h (the ECU suppresses low signals as noise), while active sensors detect speeds as low as 0.1 km/h; therefore the absence of low-speed OBD-II readings identifies a passive sensor. On three urban trajectories, the passive sensor's error variance against a tactical-grade NovAtel integrated reference was well described by a two-term exponential in speed, with variance largest at low speeds. The same dataset shows that a stereo visual odometry routine with semantic segmentation-based outlier rejection yields lower speed RMSE than the passive WSS in the 0–4 km/h range, while degrading in static, low-traffic scenes, which motivates a speed-dependent fusion strategy.

Load-bearing premise

The classification rule assumes that no passive sensor ever reports below 3 km/h and that every active sensor reports down to 0.1 km/h, so a single drive's absence of low-speed OBD-II readings on one minivan is treated as proof of a passive sensor.

Editorial extensions

If this is right

  • An integrator with only OBD-II logs can classify the wheel speed sensor type by checking whether speed values below 3 km/h ever appear.
  • For passive-WSS vehicles, speed error variance is predictable from speed alone, so a fusion filter can schedule sensor weights by the fitted curve.
  • Below roughly 4 km/h, stereo visual odometry with semantic outlier rejection offers lower speed RMSE than the passive OBD-II sensor, while OBD-II is more reliable in static-object scenes.
  • A fusion engine combining OBD-II-derived speed, stereo visual odometry, and additional classification aids should improve forward-speed estimates in GNSS-denied environments.
  • Extending the identification and variance-fitting procedure to vehicles with active wheel speed sensors would move toward the plug-and-play fusion the paper motivates.

Reading between the lines

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

  • The 3 km/h threshold may be an ECU or OEM configuration rather than a universal sensor property, so the same absence-based heuristic could be tested on other makes and models to see whether the cutoff is truly portable.
  • Because the fitted variance grows sharply at low speeds, an immediate testable extension is to inject this speed-dependent covariance into a reduced inertial sensor system and measure dead-reckoning error over repeated GNSS outages.
  • The one-vehicle validation leaves open that the exponential coefficients are vehicle-specific; fitting the same two-term form on a different passive-WSS vehicle would show whether coefficients cluster by sensor family or by tire and ECU settings.
  • The low-speed gap that passive sensors leave could be filled by direct CAN-bus wheel tick access or by RADAR ego-velocity estimation, both of which the paper cites as promising alternatives for low-speed maneuvers.
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

4 major / 7 minor

Summary. The manuscript surveys automotive speed-sensing technologies, distinguishing OBD-II from CAN-bus access, transmission-RPM-based (VSS) from ABS-wheel-speed-based (WSS) derivation, and passive from active WSS, and it analyzes three urban road trajectories collected from a single Toyota Sienna (Toronto T19.1/T19.2 and Kingston K19.1). Its two central technical claims are: (i) a sensor-identification rule stating that absence of OBD-II-reported speeds below 3 km/h identifies a passive WSS (Section IV-B1), and (ii) an empirical error-variance model for the passive WSS, sigma^2(v_f) = 0.3095 exp(-0.1241 v_f) + 0.1477 exp(-0.0009149 v_f) (Eq. 5). The paper also compares OBD-II-derived speed against two stereo visual-odometry variants (Fig. 10) to argue for multi-sensor fusion. The paper is transparent that universal applicability requires data from a broader set of vehicles, including active-WSS vehicles (Section V).

Significance. If the claims are supported, the paper would give navigation integrators a low-cost way to (a) determine the type of wheel-speed sensor from OBD-II logs alone and (b) attach a speed-dependent error variance to the OBD-II speed channel, both directly useful for RISS/INS fusion and dead-reckoning. The paper's strengths are the clear taxonomy of speed-derivation methods (Tables I-II), the public release of supporting material, and the honesty with which the variance model is labeled as an empirical fit rather than a derivation; the classification rule is falsifiable and grounded in a citable automotive behavior (Ref. [17]) rather than in the paper's own output, so there is no circularity. However, the evidence base is a single vehicle, the identification rule is tested only in the absence direction, and none of the empirical quantities carry uncertainty quantification; as written, the abstract's claim of proposing a method to identify sensor type outruns the demonstrated support.

major comments (4)
  1. [Section IV-B1 (with Section III-C)] The sensor-identification rule is load-bearing for the abstract's central claim ('we propose a method to identify the type of speed sensor in a vehicle'), but the manuscript supports only one direction of the rule, on a single Toyota Sienna: absence of low-speed OBD-II readings is observed, while the claim that active WSS always report down to 0.1 km/h is only cited from Ref. [17] and never tested. Because the rule operates on OBD-II PID 0x0D output, it is subject to OEM-specific ECU behavior, a risk the paper itself acknowledges at the start of Section III; consequently the false-positive rate of the 'no readings below 3 km/h implies passive' classification is unknown. Section V concedes the gap ('would require collecting OBD-II-derived speed data across a broad spectrum of vehicles, including those equipped with active wheel speed sensors'), and the stress-test concern therefore lands. In addition, the manuscript does not state what PID 0x0D returns when the ABS module suppresses a sub-threshold signal (zero, drop-out, or stale value), which makes the criterion and Fig. 7 ambiguous. The revision should either test the rule on vehicles with known active WSS (for example, a public CAN dataset such as comma2k19 [3]) or explicitly reframe the identification claim as a single-vehicle case study with a tentative heuristic.
  2. [Eq. (5) and Fig. 8] The error-variance model is presented as available 'for further analysis or predictions,' but as reported it is an unvalidated fit: no held-out trajectory is used, no goodness-of-fit statistic is given, and the claimed 95% confidence bounds for a, b, c, and d are never reported. The fit also covers only speed bins from 4 km/h upward, leaving the 0-4 km/h band—the regime in which the paper argues passive WSS error is largest (Figs. 9 and 10) and the regime in which the identification heuristic operates—outside the model. Please report the coefficient bounds and residuals, add a per-trajectory cross-validation, state explicitly that v_f is the reference-speed bin variable, and explain what the OBD-II port reports in the sub-threshold region so that the low-speed gap is justified.
  3. [Section IV-B1 (ground-truth reference)] All variance estimates used in Eq. (5) and the RMSE values in Fig. 10 are computed against the NovAtel integrated navigation solution, but the manuscript gives no quality metric for that reference in the urban GNSS-challenged segments (for example, PDOP, satellite count, estimated velocity uncertainty, or outage durations). If the reference drifts or is smoothed during GNSS outages, the estimated PWSS error variance is biased upward. Adding per-trajectory reference-quality statistics or a short sensitivity discussion would materially strengthen the variance model.
  4. [Fig. 10] The comparison of RMSE across speed intervals is presented as showing 'significant variations,' but no error bars, per-interval sample counts, or statistical tests are provided; since Fig. 9 shows that interval occupancy differs strongly across the three trajectories, the pooled RMSE comparison may reflect trajectory speed distributions rather than sensor error characteristics. Please report per-interval sample sizes and confidence intervals, and remove the stray numeric labels in the figure (for example, '360.' and '871.'), which appear to be rendering artifacts.
minor comments (7)
  1. [Eq. (1)] The symbol '⌊·⌉' is described as 'truncating to the nearest integer,' which conflates truncation with nearest-integer rounding; please adopt a consistent, standard notation across Eqs. (1), (2), and (4).
  2. [Fig. 8] The typeset axis labels and legend are mangled ('4-66-88-10', '7<2PWSS : Meanofvariances'); please provide a clean figure and define the bin edges precisely, including whether the '<2' bin enters the fit.
  3. [Section III-C] Because the 3 km/h and 0.1 km/h thresholds are load-bearing for the classification rule, please cite the specific pages or sections of Ref. [17] that state them rather than citing the book as a whole.
  4. [Section II] The statement that OBD-II is 'standardized and interoperable' across all OBD-II-compliant vehicles is in tension with the paper's own demonstration that low-speed reporting behavior is vehicle-specific; please qualify the claim.
  5. [Table II] Listing 'more susceptible to electromagnetic interference' as a drawback of active WSS conflicts with the usual engineering understanding that active sensors provide better signal integrity; please verify this against the cited source.
  6. [Section IV-B1] The sentence explaining the 95% confidence bounds ('there is a 95% probability that the actual values of the coefficients fall within the bounds') misstates the frequentist interpretation of a confidence interval; please rephrase or report the numeric bounds.
  7. [Eqs. (2)-(3)] The overview does not discuss how tire radius variation (inflation pressure, wear, load) affects OBD-II-derived speed, despite r_w appearing in Eqs. (2) and (3); one sentence on this known error source would improve the survey's completeness.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the error model is an explicit empirical fit and the sensor classification heuristic rests on an external cited behavior, not on the paper's own outputs.

full rationale

The central error variance model in Eq. (5) is explicitly presented as a curve fit to measured data: the paper states that a 2nd degree exponential function was selected in the data fitting process and its coefficients were determined using an optimization algorithm. This is a fit, not a first-principles derivation, and the paper does not relabel the fitted curve as an independent prediction. The sensor identification step in Section IV-B1 applies a threshold ('If such readings were present, the sensor was classified as active; otherwise, it was considered passive') that is justified by citing known passive WSS behavior in external reference [17], not by the paper's own results or equations. No equation makes the classification or the variance model equivalent to its inputs by construction. The self-citations [18] and [19] supply dataset details and a Stereo VO comparison baseline; they are not load-bearing for the central claim. The lack of validation on a vehicle with a known active sensor is a generalizability and correctness concern, not a circularity, so it does not raise the circularity score.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claims rest on four fitted curve coefficients, a hand-chosen classification threshold, and several domain assumptions about OBD-II behavior and reference accuracy. No invented entities are introduced.

free parameters (5)
  • a = 0.3095 (km/h)^2
    Amplitude of the fast-decaying exponential term in Eq. (5), fit to binned variance data from three trajectories.
  • b = -0.1241 (km/h)^-1
    Decay rate of the first exponential term in Eq. (5), fit to binned variance data.
  • c = 0.1477 (km/h)^2
    Amplitude of the slow-decaying exponential term in Eq. (5), fit to binned variance data.
  • d = -0.0009149 (km/h)^-1
    Near-zero decay rate of the second exponential term in Eq. (5), fit to binned variance data.
  • low-speed classification threshold = 3 km/h
    Hand-chosen threshold used to classify WSS type; based on known passive sensor behavior, not estimated from the collected data.
assumptions (4)
  • domain assumption Passive WSS do not report vehicle speeds below approximately 3 km/h, while active WSS report down to 0.1 km/h.
    Section IV-B1 uses the absence of sub-3 km/h readings to classify the Sienna's sensors as passive; if this threshold is manufacturer-specific or active systems also suppress low speeds, the classification fails.
  • domain assumption OBD-II PID 0x0D vehicle speed is the integer-truncated average of the four wheel speeds for ABS-equipped vehicles.
    The paper models OBD-II speed as the average of individual wheel speeds (Eq. 4); actual OEM implementations may differ, e.g., using a minimum wheel speed or rear-axle only.
  • domain assumption The NovAtel ProPak6 tactical-grade GNSS/IMU integrated solution is an accurate ground truth in urban trajectories.
    Error variance estimates are computed against this reference; in GNSS-denied urban canyons the reference itself may carry errors, especially at low speed.
  • domain assumption OBD-II speed reports are quantized to integer km/h via rounding or truncation.
    Eq. (1) and Listing 1 rely on integer encoding of vehicle speed; the exact rounding behavior affects the error model at low speeds.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Automotive Speed Estimation: Sensor Types and Error Characteristics from OBD-II to ADAS." pith.science (2026). https://pith.science/paper/YNSQSER4

@misc{pith2026250100242,
  author       = {Pith},
  title        = {Pith review of: Automotive Speed Estimation: Sensor Types and Error Characteristics from OBD-II to ADAS},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YNSQSER4}},
  note         = {Machine review of arXiv:2501.00242}
}
read the original abstract

Modern on-road navigation systems heavily depend on integrating speed measurements with inertial navigation systems (INS) and global navigation satellite systems (GNSS). Telemetry-based applications typically source speed data from the On-Board Diagnostic II (OBD-II) system. However, the method of deriving speed, as well as the types of sensors used to measure wheel speed, differs across vehicles. These differences result in varying error characteristics that must be accounted for in navigation and autonomy applications. This paper addresses this gap by examining the diverse speed-sensing technologies employed in standard automotive systems and alternative techniques used in advanced systems designed for higher levels of autonomy, such as Advanced Driver Assistance Systems (ADAS), Autonomous Driving (AD), or surveying applications. We propose a method to identify the type of speed sensor in a vehicle and present strategies for accurately modeling its error characteristics. To validate our approach, we collected and analyzed data from three long real road trajectories conducted in urban environments in Toronto and Kingston, Ontario, Canada. The results underscore the critical role of integrating multiple sensor modalities to achieve more accurate speed estimation, thus improving automotive navigation state estimation, particularly in GNSS-denied environments.

Figures

Figures reproduced from arXiv: 2501.00242 by the authors.

Figure 1
Figure 1. Interconnections of Automotive Sensors and Systems via CAN. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Methods for estimating the speed of a land vehicle without relying on GNSS or IMU integration. [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Illustration of gear ratio. In non-ABS vehicles, where speed is derived from the engine’s VSS (see [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (9 more)
Figure 4
Figure 4. Figure 4: Vehicle Speed Sensor (VSS) [PITH_FULL_IMAGE:figures/full_fig_p003_4.png]
Figure 5
Figure 5. Figure 5: Wheel Speed Sensor (WSS). The compute module calculates the speed of each individual wheel using the formula: Vw = 2π · rw · Np Nt · ∆t , (3) where Vw denotes the linear speed of an individual wheel (m/s), Np is the number of pulses detected during the sampling interva…
Figure 6
Figure 6. Figure 6: Data logging setup. In this setup, we employed the STN1110 OBD-to-UART interpreter to interface with the vehicle’s OBD-II, enabling 2GitHub repository: https://github.com/hanymragab/plans2025-resources [PITH_FULL_IMAGE:figures/full_fig_p004_6.png]
Figure 8
Figure 8. Figure 8: Error variance analysis of passive wheel speed sensor data from three [PITH_FULL_IMAGE:figures/full_fig_p005_8.png]
Figure 7
Figure 7. Figure 7: OBD-II-reported vehicle speed (green) as compared to the reference [PITH_FULL_IMAGE:figures/full_fig_p005_7.png]
Figure 9
Figure 9. Figure 9: Histogram-based Probability Mass Function (PMF) in terms of speed [PITH_FULL_IMAGE:figures/full_fig_p005_9.png]
Figure 11
Figure 11. Figure 11: presents a comparison of speed estimates from Stereo VO, with and without the SS-OR enhancement, along￾side OBD-II-derived speeds and the NovAtel reference. It highlights the variability in Stereo VO performance across different scenarios and showcases the capability …
Figure 12
Figure 12. Figure 12: Degeneration of SS-OR: Parked or stationary vehicles misclassified [PITH_FULL_IMAGE:figures/full_fig_p006_12.png]
Figure 10
Figure 10. Figure 10: Speed RMSE comparison between PWSS and Stereo VO with and [PITH_FULL_IMAGE:figures/full_fig_p006_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

19 extracted references · 17 canonical work pages

  1. [17]

    Reif, Ed., Brakes, Brake Control and Driver Assis- tance Systems: Function, Regulation and Components

    K. Reif, Ed., Brakes, Brake Control and Driver Assis- tance Systems: Function, Regulation and Components . Wiesbaden: Springer Fachmedien Wiesbaden, 2014

    work page 2014

  2. [3]

    A Commute in Data: The comma2k19 Dataset,

    H. Schafer, E. Santana, A. Haden, and R. Biasini, “A Commute in Data: The comma2k19 Dataset,” Dec. 2018, arXiv:1812.05752 [cs]

    arXiv 2018

  3. [1]

    Autonomous vehicles and intelligent automation: Applications, chal- lenges, and opportunities,

    G. Bathla, K. Bhadane, R. K. Singh, R. Kumar, R. Alu- valu, R. Krishnamurthi, A. Kumar, R. N. Thakur, S. Basheer, and M. P. Kumar Reddy, “Autonomous vehicles and intelligent automation: Applications, chal- lenges, and opportunities,” Mobile Information Systems , vol. 2022, Jan. 2022

    work page 2022

  4. [2]

    Mercedes-Benz backs redundancy for safe conditionally automated driving,

    M.-B. Group, “Mercedes-Benz backs redundancy for safe conditionally automated driving,” Sep. 2022, - Last Accessed: 2024-12-18. Section: Innovations. [Online]. Available: https://group.mercedes-benz. com/innovation/product-innovation/autonomous-driving/ redundancy-drive-pilot.html

    work page 2022

  5. [4]

    An integrated reduced inertial sensor system - RISS / GPS for land vehicle,

    U. Iqbal, A. F. Okou, and A. Noureldin, “An integrated reduced inertial sensor system - RISS / GPS for land vehicle,” in 2008 IEEE/ION Position, Location and Nav- igation Symposium. Monterey, CA, USA: IEEE, 2008, pp. 1014–1021

    work page 2008

  6. [5]

    A High-rate, Heterogeneous Data Set From The DARPA Urban Chal- lenge,

    A. S. Huang, M. Antone, E. Olson, L. Fletcher, D. Moore, S. Teller, and J. Leonard, “A High-rate, Heterogeneous Data Set From The DARPA Urban Chal- lenge,” The International Journal of Robotics Research , vol. 29, no. 13, pp. 1595–1601, Nov. 2010

    work page 2010

  7. [6]

    Ford Campus vision and lidar data set,

    G. Pandey, J. R. McBride, and R. M. Eustice, “Ford Campus vision and lidar data set,” International Journal of Robotics Research , p. 1543–1552, Nov. 2011

    work page 2011

  8. [7]

    Complex urban dataset with multi-level sensors from highly diverse urban environments,

    J. Jeong, Y . Cho, Y .-S. Shin, H. Roh, and A. Kim, “Complex urban dataset with multi-level sensors from highly diverse urban environments,” The International Journal of Robotics Research , vol. 38, no. 6, pp. 642– 657, May 2019

    work page 2019

Show all 19 references
  1. [8]

    A Multi-sensor Traffic Scene Dataset with Omnidirectional Video,

    P. Koschorrek, T. Piccini, P. ¨Oberg, M. Felsberg, L. Nielsen, and R. Mester, “A Multi-sensor Traffic Scene Dataset with Omnidirectional Video,” in 2013 IEEE Conference on Computer Vision and Pattern Recognition Workshops, Jun. 2013, pp. 727–734, ISSN: 2160-7516

    2013

  2. [9]

    Gnss/multisensor fusion using continuous-time factor graph optimization for robust localization,

    H. Zhang, C.-C. Chen, H. Vallery, and T. D. Barfoot, “Gnss/multisensor fusion using continuous-time factor graph optimization for robust localization,” IEEE Trans- actions on Robotics , vol. 40, pp. 4003–4023, 2024

    2024

  3. [10]

    nuScenes: A Multimodal Dataset for Autonomous Driv- ing,

    H. Caesar, V . Bankiti, A. H. Lang, S. V ora, V . E. Liong, Q. Xu, A. Krishnan, Y . Pan, G. Baldan, and O. Beijbom, “nuScenes: A Multimodal Dataset for Autonomous Driv- ing,” in 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) . Seattle, W A, USA: IEE...

    2020

  4. [11]

    A2D2: Audi Autonomous Driving Dataset,

    J. Geyer, Y . Kassahun, M. Mahmudi, X. Ricou, R. Durgesh, A. S. Chung, L. Hauswald, V . H. Pham, M. M ¨uhlegg, S. Dorn, T. Fernandez, M. J ¨anicke, S. Mi- rashi, C. Savani, M. Sturm, O. V orobiov, M. Oelker, S. Garreis, and P. Schuberth, “A2D2: Audi Autonomous Driving Dataset,...

    2020 arXiv

  5. [12]

    Noureldin, T

    A. Noureldin, T. B. Karamat, and J. Georgy, Fundamen- tals of Inertial Navigation, Satellite-based Positioning and their Integration . Berlin, Heidelberg: Springer, 2013, ISBN: 978-3-642-30466

    2013

  6. [13]

    GNSS/IMU/ODO/LiDAR-SLAM Integrated Navigation System Using IMU/ODO Pre-Integration,

    L. Chang, X. Niu, and T. Liu, “GNSS/IMU/ODO/LiDAR-SLAM Integrated Navigation System Using IMU/ODO Pre-Integration,” Sensors, vol. 20, no. 17, p. 4702, Jan. 2020, Multidisciplinary Digital Publishing Institute

    2020

  7. [14]

    An Enhanced Error Model for EKF-Based Tightly-Coupled Integration of GPS and Land Vehicle’s Motion Sensors,

    T. B. Karamat, M. M. Atia, and A. Noureldin, “An Enhanced Error Model for EKF-Based Tightly-Coupled Integration of GPS and Land Vehicle’s Motion Sensors,” Sensors, vol. 15, no. 9, pp. 24 269–24 296, Sep. 2015, Multidisciplinary Digital Publishing Institute

    2015

  8. [15]

    Stereo Vision-Based Seman- tic 3D Object and Ego-Motion Tracking for Autonomous Driving,

    P. Li, T. Qin, and S. Shen, “Stereo Vision-Based Seman- tic 3D Object and Ego-Motion Tracking for Autonomous Driving,” vol. 11206, pp. 664–679, 2018, Lecture Notes in Computer Science

    2018

  9. [16]

    To- ward Land Vehicle Ego-Velocity Estimation Using Deep Learning and Automotive Radars,

    P. R. M. de Araujo, A. Noureldin, and S. Givigi, “To- ward Land Vehicle Ego-Velocity Estimation Using Deep Learning and Automotive Radars,” IEEE Transactions on Radar Systems, vol. 2, pp. 460–470, 2024

    2024

  10. [18]

    Integration of Topological Maps with GNSS and Onboard Sensors for Robust Land Vehicle Navigation,

    H. Ragab, “Integration of Topological Maps with GNSS and Onboard Sensors for Robust Land Vehicle Navigation,” Ph.D. dissertation, Queen’s University, Kingston, Ontario, Canada, Jul. 2024. [Online]. Available: https://hdl.handle.net/1974/33163

    2024

  11. [19]

    The Utilization of DNN-based Semantic Segmenta- tion for Improving Low-Cost Integrated Stereo Visual Odometry in Challenging Urban Environments,

    H. Ragab, M. Elhabiby, S. Givigi, and A. Noureldin, “The Utilization of DNN-based Semantic Segmenta- tion for Improving Low-Cost Integrated Stereo Visual Odometry in Challenging Urban Environments,” in 2020 IEEE/ION Position, Location and Navigation Symposium (PLANS). Portland...

    2020

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.