Belief-Space Residual Risk for Automated Driving under Localization Uncertainty

Frank Diermeyer; Nijinshan Karunainayagam; Nils Gehrke

arxiv: 2605.12710 · v1 · pith:DBN6PDFNnew · submitted 2026-05-12 · 💻 cs.RO

Belief-Space Residual Risk for Automated Driving under Localization Uncertainty

Nijinshan Karunainayagam , Nils Gehrke , Frank Diermeyer This is my paper

Pith reviewed 2026-05-14 20:00 UTC · model grok-4.3

classification 💻 cs.RO

keywords automated drivingresidual risklocalization uncertaintybelief spaceGaussian distributioncollision probabilityparticle-based estimationsafety assessment

0 comments

The pith

Residual risk assessment for automated driving is extended into belief space by modeling ego pose uncertainty as a Gaussian distribution and reformulating risk as an expectation over that belief.

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

The paper shows how to move residual risk calculations from a fixed ego pose to a distribution over possible poses. Existing methods treat the vehicle's location as known exactly and focus on errors in detecting other objects. By treating pose as a Gaussian belief and fusing its covariance with object uncertainties inside a particle-based estimator, the approach computes expected collision probabilities under localization error. This matters because real vehicles in cities face substantial pose uncertainty from GPS, maps, and weather, which can hide or exaggerate safety threats. The reformulation turns degradation-induced risk into an average taken over the belief distribution rather than a single point calculation.

Core claim

Residual risk is reformulated as the expected degradation-induced risk over the ego pose belief distribution, with localization uncertainty incorporated through covariance fusion of ego and object uncertainties within a particle-based risk estimation framework.

What carries the argument

Belief-space residual risk computed via Gaussian ego pose modeling and covariance fusion to obtain collision probabilities over the pose belief distribution.

If this is right

Safety metrics become sensitive to localization quality in addition to perception errors, allowing planners to trade off map accuracy against risk thresholds.
Particle-based estimators can now propagate pose uncertainty into risk without separate Monte Carlo sampling over poses.
Degradation functions that penalize proximity to obstacles will produce higher expected risk when pose covariance is large.
Urban and adverse-weather scenarios receive quantitatively higher residual risk scores than highway scenarios with the same object configuration.

Where Pith is reading between the lines

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

The same Gaussian-fusion step could be applied to other uncertainty sources such as map mismatch or sensor calibration drift if they are also modeled as additive covariances.
A planner using this risk could reduce speed or increase clearance margins proportionally to current localization covariance, producing behavior that adapts to changing GPS conditions.
Extending the formulation to non-Gaussian pose beliefs would require replacing the covariance fusion step with a more general expectation over particles or mixture models.

Load-bearing premise

Ego pose uncertainty is adequately captured by a single Gaussian distribution whose covariance can be fused directly with object uncertainties to give accurate collision probabilities.

What would settle it

A controlled test where the vehicle is driven through a known urban scene with recorded ground-truth localization error; if the belief-space risk values do not match observed near-miss frequencies better than the deterministic baseline, the extension does not hold.

Figures

Figures reproduced from arXiv: 2605.12710 by Frank Diermeyer, Nijinshan Karunainayagam, Nils Gehrke.

**Figure 2.** Figure 2: Visualization of the TUMDOT-Dataset with the [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Figure 3: Spatial heat maps of the mean residual risk of the selected scenarios. The nominal ego positions are each marked [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

**Figure 4.** Figure 4: Behavior of the mean residual risk with increasing [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗

**Figure 5.** Figure 5: Risk variability over increasing localization uncer [PITH_FULL_IMAGE:figures/full_fig_p005_5.png] view at source ↗

**Figure 6.** Figure 6: Exceedance Probability over increasing localization [PITH_FULL_IMAGE:figures/full_fig_p006_6.png] view at source ↗

read the original abstract

Residual risk metrics have recently been introduced to assess the safety implications of automated driving systems. Existing approaches typically assume a deterministic ego pose and concentrate mainly on perception errors related to surrounding objects and latency effects. In practice, however, automated vehicles operate under considerable localization uncertainty, especially in complex urban settings and in adverse weather conditions. This work extends the spatial residual risk formulation to the belief space by explicitly modeling ego pose uncertainty as a Gaussian distribution. Residual risk is reformulated as the expected degradation-induced risk over the ego pose belief distribution. Within a particle-based risk estimation framework, localization uncertainty is incorporated into the computation of collision probabilities through covariance fusion of ego and object uncertainties.

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.

Desk Editor's Note private letter to a colleague

This paper extends residual risk to belief space by modeling ego pose as Gaussian and using covariance fusion in particles, but the Gaussian assumption and linearity for collision probs need checking against real localization errors.

read the letter

The main point is a straightforward extension: residual risk gets reformulated as the expected value of degradation risk over a Gaussian belief on ego pose, with localization uncertainty folded in via covariance fusion inside the existing particle estimator. This directly addresses the gap where prior residual risk work assumed a fixed ego position, which matters for urban driving and weather where localization drifts. The reformulation itself is clean and stays within standard Gaussian assumptions without adding free parameters or self-referential fits. It builds logically on the cited spatial residual risk literature by making the expectation explicit over the belief distribution. That part is useful for safety assessment pipelines that already use residual risk. The soft spot is the reliance on the fused covariance staying representative. When the underlying risk function has sharp thresholds near zero clearance, first- and second-order moments alone can miss the probability mass that actually crosses into collision; real localization errors are often multimodal rather than neatly Gaussian. The abstract gives no derivations or error bounds, so the full paper needs to show whether the particle framework compensates or if the bias stays small in practice. This is for engineers and researchers who validate automated driving safety metrics and already work with residual risk tools. A reader looking for incremental improvements to existing frameworks will get value from the explicit belief-space step. It deserves peer review because the core idea is well-motivated and the method is reproducible in principle, even if the validation details will decide how far it travels.

Referee Report

2 major / 2 minor

Summary. The manuscript extends spatial residual risk metrics to belief space for automated driving safety assessment. It models ego pose uncertainty as a Gaussian distribution, reformulates residual risk as the expected degradation-induced risk over the ego-pose belief, and incorporates localization uncertainty into collision probabilities via covariance fusion of ego and object uncertainties inside a particle-based estimator. The approach targets urban and adverse-weather scenarios where deterministic-pose assumptions are unrealistic.

Significance. If the Gaussian belief and covariance-fusion steps are shown to be accurate, the work would fill a recognized gap between perception-focused residual risk and full localization uncertainty, potentially enabling tighter safety bounds for automated vehicles. The parameter-free character of the reformulation (no new fitted parameters introduced) and the particle-based implementation are concrete strengths that support reproducibility and falsifiability.

major comments (2)

[Method] The central claim that covariance fusion inside the particle estimator yields correct collision probabilities rests on the implicit assumption that the fused uncertainty remains Gaussian and that the residual-risk function is sufficiently linear. When the risk function contains sharp thresholds (near-zero clearance), first- and second-order moments alone cannot recover the expectation; this is a load-bearing issue for the safety claims. A concrete error bound or Monte-Carlo validation against non-Gaussian localization errors is required (Method section, covariance-fusion paragraph).
[Section 3] No derivation or error analysis is supplied for the expectation of the degradation-induced risk over the Gaussian belief distribution. The abstract states the reformulation but the manuscript must show whether the particle estimator converges to the true integral or merely approximates it under the Gaussian assumption (Section 3, Eq. defining the belief-space residual risk).

minor comments (2)

[Method] Notation for the fused covariance matrix is introduced without an explicit equation number; add a numbered definition to avoid ambiguity when readers compare with standard Kalman fusion.
[Method] The particle-based framework description would benefit from a small pseudocode block or diagram showing how ego-pose samples are drawn and fused with object covariances.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback on our manuscript. We address each major comment below and describe the revisions we will make to strengthen the presentation and validation of our approach.

read point-by-point responses

Referee: [Method] The central claim that covariance fusion inside the particle estimator yields correct collision probabilities rests on the implicit assumption that the fused uncertainty remains Gaussian and that the residual-risk function is sufficiently linear. When the risk function contains sharp thresholds (near-zero clearance), first- and second-order moments alone cannot recover the expectation; this is a load-bearing issue for the safety claims. A concrete error bound or Monte-Carlo validation against non-Gaussian localization errors is required (Method section, covariance-fusion paragraph).

Authors: We agree that the covariance-fusion step implicitly relies on a Gaussian approximation for the combined ego-object uncertainty and that this may not fully capture expectations for highly non-linear risk functions with sharp thresholds. In the revised manuscript we will augment the Method section with a dedicated Monte-Carlo validation study. This study will compare the particle-based estimator against direct sampling from non-Gaussian localization error models (e.g., heavy-tailed or multimodal distributions) in near-zero-clearance scenarios, and will report empirical error bounds on the resulting collision probabilities. revision: yes
Referee: [Section 3] No derivation or error analysis is supplied for the expectation of the degradation-induced risk over the Gaussian belief distribution. The abstract states the reformulation but the manuscript must show whether the particle estimator converges to the true integral or merely approximates it under the Gaussian assumption (Section 3, Eq. defining the belief-space residual risk).

Authors: We acknowledge the absence of an explicit derivation and convergence analysis for the belief-space expectation. In the revision we will expand Section 3 to include (i) a step-by-step derivation of the expectation of the degradation-induced risk over the Gaussian ego-pose belief and (ii) an analysis of the particle estimator’s convergence to the true integral under the stated Gaussian assumption, together with a discussion of the conditions under which the approximation becomes exact. revision: yes

Circularity Check

0 steps flagged

Derivation self-contained; no reductions to fitted inputs or self-citations

full rationale

The paper extends an existing spatial residual risk formulation to belief space by modeling ego pose uncertainty as a Gaussian and defining residual risk as the expectation of degradation-induced risk over that belief distribution, with localization uncertainty incorporated via covariance fusion inside a particle-based estimator. No equations or definitions in the provided text reduce the new risk quantity to a fitted parameter, a renamed input, or a load-bearing self-citation by construction. The reformulation rests on external residual risk literature and standard Gaussian/probabilistic assumptions that are independent of the target result, so the central claim remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the modeling choice that localization uncertainty is Gaussian and that covariance fusion correctly combines uncertainties for collision probability; no free parameters or new entities are introduced in the abstract.

axioms (1)

domain assumption Ego pose uncertainty can be modeled as a Gaussian distribution
Stated directly in the abstract as the basis for extending residual risk to belief space.

pith-pipeline@v0.9.0 · 5409 in / 1172 out tokens · 62896 ms · 2026-05-14T20:00:51.804121+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

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

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

Residual risk is reformulated as the expected degradation-induced risk over the ego pose belief distribution... covariance fusion of ego and object uncertainties
IndisputableMonolith/Foundation/AbsoluteFloorClosure.lean absolute_floor_iff_bare_distinguishability unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

xe ∼ N(ˆxe, Σe) ... Σrel = Σo + Σe

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

23 extracted references · 23 canonical work pages

[1]

Road vehicles – Functional safety,

ISO, “Road vehicles – Functional safety,” 2011

work page 2011
[2]

Is it safe to drive? an overview of factors, metrics, and datasets for driveability assessment in au- tonomous driving,

J. Guo, U. Kurup, and M. Shah, “Is it safe to drive? an overview of factors, metrics, and datasets for driveability assessment in au- tonomous driving,”IEEE Transactions on Intelligent Transportation Systems, vol. 21, no. 8, pp. 3135–3151, 2020

work page 2020
[3]

State Estimation Based on Generalized Gaussian Distributions,

X. Li and Y . Xie, “State Estimation Based on Generalized Gaussian Distributions,”Metrology and Measurement Systems, vol. 20, no. 1, pp. 65–76, 2013

work page 2013
[4]

Tumdot–muc: Data collection and processing of multimodal trajectories collected by aerial drones,

A. Kutsch, M. Margreiter, and K. Bogenberger, “Tumdot–muc: Data collection and processing of multimodal trajectories collected by aerial drones,”Data Science for Transportation, vol. 6, no. 2, Aug. 2024, publisher Copyright: © The Author(s) 2024

work page 2024
[5]

Introducing spatial residual risk for information degradation in automated driving,

N. Gehrke and F. Diermeyer, “Introducing spatial residual risk for information degradation in automated driving,” inIV 2025 - 36th IEEE Intelligent V ehicles Symposium, ser. IEEE Intelligent Vehicles Symposium, Proceedings. Institute of Electrical and Electronics En- gineers Inc., 2025, pp. 821–827, publisher Copyright: © 2025 IEEE.; 36th IEEE Intelligent...

work page 2025
[6]

Risk assessment in autonomous driving: a comprehensive survey of risk sources, methodologies, and system architectures,

D. Lu, H. Du, Z. Wu, and S. Yang, “Risk assessment in autonomous driving: a comprehensive survey of risk sources, methodologies, and system architectures,”Autonomous Intelligent Systems, vol. 5, 2025. [Online]. Available: https://api.semanticscholar.org/CorpusID:281518313

work page 2025
[7]

Risk assessment methodologies for autonomous driving: A survey,

W. M. D. Chia, S. L. Keoh, C. Goh, and C. Johnson, “Risk assessment methodologies for autonomous driving: A survey,”IEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 10, pp. 16 923– 16 939, 2022

work page 2022
[8]

R. J. Mikulak, R. McDermott, and M. Beauregard,The basics of FMEA. CRC press, 2017

work page 2017
[9]

Fault tree handbook,

W. E. Vesely, F. F. Goldberg, N. H. Roberts, and D. F. Haasl, “Fault tree handbook,” Tech. Rep., 1981

work page 1981
[10]

Stpa: A new hazard analysis technique,

N. G. Leveson, “Stpa: A new hazard analysis technique,” inEngineering a Safer World: Systems Thinking Applied to Safety. The MIT Press, 01 2012. [Online]. Available: https://doi.org/10.7551/mitpress/8179.003.0013

work page doi:10.7551/mitpress/8179.003.0013 2012
[11]

Taxonomy and definitions for terms related to driving automa- tion systems for on-road motor vehicles,

SAE, “Taxonomy and definitions for terms related to driving automa- tion systems for on-road motor vehicles,” 2021

work page 2021
[12]

Cooperative driving of connected autonomous vehicles using responsibility-sensitive safety (rss) rules,

M. Khayatian, M. Mehrabian, H. Allamsetti, K.-W. Liu, P.-Y . Huang, C.-W. Lin, and A. Shrivastava, “Cooperative driving of connected autonomous vehicles using responsibility-sensitive safety (rss) rules,” inProceedings of the ACM/IEEE 12th International Conference on Cyber-Physical Systems, ser. ICCPS ’21. New York, NY , USA: Association for Computing Mac...

work page doi:10.1145/3450267.3450530 2021
[13]

A new integrated collision risk assessment methodology for autonomous vehicles,

C. Katrakazas, M. Quddus, and W.-H. Chen, “A new integrated collision risk assessment methodology for autonomous vehicles,”Accident Analysis & Prevention, vol. 127, pp. 61–79, 2019. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S0001457518306614

work page 2019
[14]

Deep predictive models for collision risk assessment in autonomous driving,

M. Strickland, G. Fainekos, and H. B. Amor, “Deep predictive models for collision risk assessment in autonomous driving,” in2018 IEEE International Conference on Robotics and Automation (ICRA), 2018, pp. 4685–4692

work page 2018
[15]

In: Principles of Modeling - Essays Dedicated to Edward A

M. T ¨orngren and U. Sellgren, “Complexity challenges in development of cyber-physical systems,” inPrinciples of Modeling, M. Lohstroh, P. Derler, and M. Sirjani, Eds. Cham: Springer International Publishing, 2018, pp. 478–503. [Online]. Available: https://doi.org/10.1007/978-3-319-95246-8 27

work page doi:10.1007/978-3-319-95246-8 2018
[16]

The road to safety: A review of uncertainty and applications to autonomous driving perception,

B. Ara ´ujo, J. F. Teixeira, J. Fonseca, R. Cerqueira, and S. C. Beco, “The road to safety: A review of uncertainty and applications to autonomous driving perception,”Entropy, vol. 26, no. 8, 2024. [Online]. Available: https://www.mdpi.com/1099-4300/26/8/634

work page 2024
[17]

Safety challenges for autonomous vehicles in the absence of connectivity,

A. Shetty, M. Yu, A. Kurzhanskiy, O. Grembek, H. Tavafoghi, and P. Varaiya, “Safety challenges for autonomous vehicles in the absence of connectivity,”Transportation Research Part C: Emerging Technologies, vol. 128, p. 103133, 2021. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S0968090X21001522

work page 2021
[18]

A comprehensive review on limitations of autonomous driving and its impact on accidents and collisions,

A. Chougule, V . Chamola, A. Sam, F. R. Yu, and B. Sikdar, “A comprehensive review on limitations of autonomous driving and its impact on accidents and collisions,”IEEE Open Journal of V ehicular Technology, vol. 5, pp. 142–161, 2024

work page 2024
[19]

Probabilistic collision checking with chance constraints,

N. E. Du Toit and J. W. Burdick, “Probabilistic collision checking with chance constraints,”IEEE Transactions on Robotics, vol. 27, no. 4, pp. 809–815, 2011

work page 2011
[20]

Risk-aware motion planning and control using cvar-constrained optimization,

A. Hakobyan, G. C. Kim, and I. Yang, “Risk-aware motion planning and control using cvar-constrained optimization,”IEEE Robotics and Automation Letters, vol. 4, no. 4, pp. 3924–3931, 2019

work page 2019
[21]

Fast collision probability estimation for automated driving using multi- circular shape approximations,

L. Tolksdorf, C. Birkner, A. Tejada, and N.{Van De Wouw}, “Fast collision probability estimation for automated driving using multi- circular shape approximations,” in35th IEEE Intelligent V ehicles Symposium, IV 2024. United States: Institute of Electrical and Electronics Engineers, Jul. 2024, pp. 2529–2536, 35th IEEE Intel- ligent Vehicles Symposium, IV ...

work page 2024
[22]

Precise and efficient collision prediction under uncertainty in autonomous driving,

M. Kaufeld and J. Betz, “Precise and efficient collision prediction under uncertainty in autonomous driving,” 10 2025

work page 2025
[23]

Risk contours map for risk bounded motion planning under perception uncertainties,

B. C. W. Ashkan M. Jasour, “Risk contours map for risk bounded motion planning under perception uncertainties,” Robotics: Science and Systems, 2019. [Online]. Available: https://www.roboticsproceedings.org/rss15/

work page 2019

[1] [1]

Road vehicles – Functional safety,

ISO, “Road vehicles – Functional safety,” 2011

work page 2011

[2] [2]

Is it safe to drive? an overview of factors, metrics, and datasets for driveability assessment in au- tonomous driving,

J. Guo, U. Kurup, and M. Shah, “Is it safe to drive? an overview of factors, metrics, and datasets for driveability assessment in au- tonomous driving,”IEEE Transactions on Intelligent Transportation Systems, vol. 21, no. 8, pp. 3135–3151, 2020

work page 2020

[3] [3]

State Estimation Based on Generalized Gaussian Distributions,

X. Li and Y . Xie, “State Estimation Based on Generalized Gaussian Distributions,”Metrology and Measurement Systems, vol. 20, no. 1, pp. 65–76, 2013

work page 2013

[4] [4]

Tumdot–muc: Data collection and processing of multimodal trajectories collected by aerial drones,

A. Kutsch, M. Margreiter, and K. Bogenberger, “Tumdot–muc: Data collection and processing of multimodal trajectories collected by aerial drones,”Data Science for Transportation, vol. 6, no. 2, Aug. 2024, publisher Copyright: © The Author(s) 2024

work page 2024

[5] [5]

Introducing spatial residual risk for information degradation in automated driving,

N. Gehrke and F. Diermeyer, “Introducing spatial residual risk for information degradation in automated driving,” inIV 2025 - 36th IEEE Intelligent V ehicles Symposium, ser. IEEE Intelligent Vehicles Symposium, Proceedings. Institute of Electrical and Electronics En- gineers Inc., 2025, pp. 821–827, publisher Copyright: © 2025 IEEE.; 36th IEEE Intelligent...

work page 2025

[6] [6]

Risk assessment in autonomous driving: a comprehensive survey of risk sources, methodologies, and system architectures,

D. Lu, H. Du, Z. Wu, and S. Yang, “Risk assessment in autonomous driving: a comprehensive survey of risk sources, methodologies, and system architectures,”Autonomous Intelligent Systems, vol. 5, 2025. [Online]. Available: https://api.semanticscholar.org/CorpusID:281518313

work page 2025

[7] [7]

Risk assessment methodologies for autonomous driving: A survey,

W. M. D. Chia, S. L. Keoh, C. Goh, and C. Johnson, “Risk assessment methodologies for autonomous driving: A survey,”IEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 10, pp. 16 923– 16 939, 2022

work page 2022

[8] [8]

R. J. Mikulak, R. McDermott, and M. Beauregard,The basics of FMEA. CRC press, 2017

work page 2017

[9] [9]

Fault tree handbook,

W. E. Vesely, F. F. Goldberg, N. H. Roberts, and D. F. Haasl, “Fault tree handbook,” Tech. Rep., 1981

work page 1981

[10] [10]

Stpa: A new hazard analysis technique,

N. G. Leveson, “Stpa: A new hazard analysis technique,” inEngineering a Safer World: Systems Thinking Applied to Safety. The MIT Press, 01 2012. [Online]. Available: https://doi.org/10.7551/mitpress/8179.003.0013

work page doi:10.7551/mitpress/8179.003.0013 2012

[11] [11]

Taxonomy and definitions for terms related to driving automa- tion systems for on-road motor vehicles,

SAE, “Taxonomy and definitions for terms related to driving automa- tion systems for on-road motor vehicles,” 2021

work page 2021

[12] [12]

Cooperative driving of connected autonomous vehicles using responsibility-sensitive safety (rss) rules,

M. Khayatian, M. Mehrabian, H. Allamsetti, K.-W. Liu, P.-Y . Huang, C.-W. Lin, and A. Shrivastava, “Cooperative driving of connected autonomous vehicles using responsibility-sensitive safety (rss) rules,” inProceedings of the ACM/IEEE 12th International Conference on Cyber-Physical Systems, ser. ICCPS ’21. New York, NY , USA: Association for Computing Mac...

work page doi:10.1145/3450267.3450530 2021

[13] [13]

A new integrated collision risk assessment methodology for autonomous vehicles,

C. Katrakazas, M. Quddus, and W.-H. Chen, “A new integrated collision risk assessment methodology for autonomous vehicles,”Accident Analysis & Prevention, vol. 127, pp. 61–79, 2019. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S0001457518306614

work page 2019

[14] [14]

Deep predictive models for collision risk assessment in autonomous driving,

M. Strickland, G. Fainekos, and H. B. Amor, “Deep predictive models for collision risk assessment in autonomous driving,” in2018 IEEE International Conference on Robotics and Automation (ICRA), 2018, pp. 4685–4692

work page 2018

[15] [15]

In: Principles of Modeling - Essays Dedicated to Edward A

M. T ¨orngren and U. Sellgren, “Complexity challenges in development of cyber-physical systems,” inPrinciples of Modeling, M. Lohstroh, P. Derler, and M. Sirjani, Eds. Cham: Springer International Publishing, 2018, pp. 478–503. [Online]. Available: https://doi.org/10.1007/978-3-319-95246-8 27

work page doi:10.1007/978-3-319-95246-8 2018

[16] [16]

The road to safety: A review of uncertainty and applications to autonomous driving perception,

B. Ara ´ujo, J. F. Teixeira, J. Fonseca, R. Cerqueira, and S. C. Beco, “The road to safety: A review of uncertainty and applications to autonomous driving perception,”Entropy, vol. 26, no. 8, 2024. [Online]. Available: https://www.mdpi.com/1099-4300/26/8/634

work page 2024

[17] [17]

Safety challenges for autonomous vehicles in the absence of connectivity,

A. Shetty, M. Yu, A. Kurzhanskiy, O. Grembek, H. Tavafoghi, and P. Varaiya, “Safety challenges for autonomous vehicles in the absence of connectivity,”Transportation Research Part C: Emerging Technologies, vol. 128, p. 103133, 2021. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S0968090X21001522

work page 2021

[18] [18]

A comprehensive review on limitations of autonomous driving and its impact on accidents and collisions,

A. Chougule, V . Chamola, A. Sam, F. R. Yu, and B. Sikdar, “A comprehensive review on limitations of autonomous driving and its impact on accidents and collisions,”IEEE Open Journal of V ehicular Technology, vol. 5, pp. 142–161, 2024

work page 2024

[19] [19]

Probabilistic collision checking with chance constraints,

N. E. Du Toit and J. W. Burdick, “Probabilistic collision checking with chance constraints,”IEEE Transactions on Robotics, vol. 27, no. 4, pp. 809–815, 2011

work page 2011

[20] [20]

Risk-aware motion planning and control using cvar-constrained optimization,

A. Hakobyan, G. C. Kim, and I. Yang, “Risk-aware motion planning and control using cvar-constrained optimization,”IEEE Robotics and Automation Letters, vol. 4, no. 4, pp. 3924–3931, 2019

work page 2019

[21] [21]

Fast collision probability estimation for automated driving using multi- circular shape approximations,

L. Tolksdorf, C. Birkner, A. Tejada, and N.{Van De Wouw}, “Fast collision probability estimation for automated driving using multi- circular shape approximations,” in35th IEEE Intelligent V ehicles Symposium, IV 2024. United States: Institute of Electrical and Electronics Engineers, Jul. 2024, pp. 2529–2536, 35th IEEE Intel- ligent Vehicles Symposium, IV ...

work page 2024

[22] [22]

Precise and efficient collision prediction under uncertainty in autonomous driving,

M. Kaufeld and J. Betz, “Precise and efficient collision prediction under uncertainty in autonomous driving,” 10 2025

work page 2025

[23] [23]

Risk contours map for risk bounded motion planning under perception uncertainties,

B. C. W. Ashkan M. Jasour, “Risk contours map for risk bounded motion planning under perception uncertainties,” Robotics: Science and Systems, 2019. [Online]. Available: https://www.roboticsproceedings.org/rss15/

work page 2019