Practical validation of synthetic pre-crash scenarios

arXiv:2605.04564 · cs.RO · Submitted 2026-05-06 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Robotics Radio. Generated commentary on the latest robotics and control papers.

Rosa: I'm Rosa.

Dev: And I'm Dev, here with Taro.

Rosa: We're looking at the paper: "Practical validation of synthetic pre-crash scenarios".

Dev: Here's the gist of the paper. The paper addresses the gap in robustly evaluating synthetic pre-crash scenarios' practical equivalence to real-world counterparts by extending a Bayesian Region of Practical Equivalence (ROPE)-based framework with a binning-based approach.

Rosa: Pruning redundant reasoning tokens The study aims to assess whether synthetic pre-crash datasets are practically equivalent to their real-world counterparts for the intended safety impact assessment.

Dev: Conventional significance testing is inadequate because it focuses on detecting differences rather than establishing practical equivalence.

Rosa: The research introduces two binning-based statistics to measure practically meaningful distributional differences between datasets in the context of safety impact assessment.

Title and authors: Rosa: So, we're looking at a paper called "Practical validation of synthetic pre-crash scenarios," by Jian Wua, Ulrich Sandera, Carol Flannaganb, and Jonas Bargman. It’s about making sure the computer simulations we use for crash testing actually match what happens in the real world.

Dev: Right. The title says they're focused on practical validation of synthetic pre-crash scenarios. Basically, they’re tackling that big gap where we can't tell if our virtual data is actually similar enough to real-world data for safety assessments.

Rosa: Exactly. They’re saying conventional statistical tests aren't good enough because they just look for differences, not if the datasets are practically equivalent for what we actually need to do—which is assessing safety impact.

Taro: So, if the synthetic data is statistically different from real data but that difference isn't dangerous in practice, conventional methods miss it.

Dev: That’s right. The paper proposes extending an existing framework, the Bayesian Region of Practical Equivalence or ROPE testing, by adding a binning-based approach to define what equivalence actually means in this context.

Rosa: It sounds like they're moving away from just checking if two things are statistically different and focusing instead on whether they are similar enough for the intended assessment purpose.

Taro: That sounds useful because we often get caught when synthetic data is skewed toward certain types of crashes, and we don't catch that bias with standard methods.

The paper's summary: Dev: Okay, so the core idea here is that they introduce two binning-based statistics, which they call theta and Theta, to measure practically meaningful differences between datasets in the context of safety impact assessment.

Rosa: They are trying to move beyond just significance testing and give us quantitative assessments of practical equivalence. One statistic, theta captures the maximum weighted absolute relative deviation across bins, which they say represents a worst-case perspective on where the difference might be most severe.

Taro: And the other one, Theta, it’s an aggregate perspective; it’s like looking at the total weighted absolute deviation across all bins together. That should give you a broader picture of how far apart the distributions are in general.

Dev: They explain how these statistics work by first selecting metrics that matter most to our safety assessment and then fitting Bayesian distribution models separately to the empirical distributions of those metrics in both datasets.

Rosa: They use expert judgment to choose which Bayesian models to fit, and then they get posterior samples from those optimal models fitted to both the reference and synthetic datasets.

Taro: So, they are using these two statistics, theta and Theta, derived from these Bayesian models, to quantify the practical difference between the distributions of our chosen metrics across the two datasets.

Dev: That’s right. The goal is to see if those statistics fall within a defined region of practical equivalence that we establish based on expert input and the desired safety assessment goal.

The paper's improvements: Rosa: One major improvement they propose is using these two binning-based statistics, theta and Theta, because they give us diagnostic insight into which specific regions of the data are driving the non-equivalence.

Dev: That’s helpful because we don't just get a pass or fail; we get to see where the synthetic data is diverging most critically from what we expect in the real world.

Taro: And they incorporate bin weights that they derive empirically by re-simulating reference scenarios with a specific driving automation system, like the Automatic Emergency Braking system, to generate system-specific outcome data.

Rosa: That weighting step is key because it makes sure the equivalence testing isn't just checking general distributions; it’s focusing on conditions where our safety systems are less effective at mitigating crashes.

Dev: And they show that this approach works when you use careful weighting, like in their case study comparing a reference dataset with a PCM dataset, where the raw PCM data failed because it had a systematic bias toward more severe cases.

Taro: So, the implication is that simply having good data isn't enough; you have to be smart about how you weigh those scenarios to make sure they represent the conditions that matter most for safety.

Conclusion: Rosa: So, wrapping up this paper on "Practical validation of synthetic pre-crash scenarios," the main point is that while appropriate weighting can definitely mitigate bias in injury-focused datasets, it can't fix every structural difference or unobserved things.

Dev: They conclude that equivalence testing shouldn't just be a one-off pass or fail criterion; it should be treated as an iterative diagnostic tool to guide how we refine our synthetic datasets.

Taro: I think what’s important is the systematic workflow they give us, specifying the metrics and applying those binning statistics theta and Theta to get interpretable measures of practical equivalence between synthetic and reference data.

Rosa: And they show that their power analysis confirms that across all metrics and both statistics, the estimated power exceeds zero point eight, in most cases approaching one point zero under their settings, which shows the framework is reliably detecting practical equivalence.

Dev: So, it gives practitioners a structured way to get decision-relevant measures with practical equivalence between synthetic and reference data without relying on those older significance testing methods that just detect differences.

Taro: For future work, they suggest expanding this framework to include multivariate distribution modeling so we can account for how different metrics might be correlated with each other in our safety assessments.

Jian Wua, Ulrich Sandera, Carol Flannagan, Jonas Bargman

Safety Center, Volvo Cars · Department of Mechanics and Maritime Sciences, Chalmers University of Technology Department of Mechanics and Maritime Sciences, University of Michigan Transportation Research Institute

cs.RO

Submitted: 2026-05-06

Updated: 2026-05-06

DOI: 10.1016/j.aap.2026.108815

Code: https://github.com/JianWu09/bayes-binned-equivalence

License: http://creativecommons.org/licenses/by-nc-nd/4.0/

Importance score: 89/100

The gist: The paper addresses the gap in robustly evaluating synthetic pre-crash scenarios' practical equivalence to real-world counterparts by extending a Bayesian Region of Practical Equivalence (ROPE)-based…

Key concepts

Practical Equivalence Testing
This is a new approach to checking if two datasets (synthetic and real) are 'similar enough' for a specific safety goal. Instead of just looking for any difference, it measures how similar they are in the contexts that matter most for assessing safety impact, ensuring the synthetic data truly represents real-world risks.
Binning-Based Statistics ($\theta$ and $\Theta$)
These are two specific mathematical tools used to quantify practical differences between datasets. $\theta$ measures the worst-case deviation across bins, focusing on localized problems, while $\Theta$ captures the total weighted deviation, offering an aggregate view of how much the distributions differ overall.
Bin Weights
These are weights assigned to different data bins based on re-simulations using a specific system (like an Automatic Emergency Braking system). These weights prioritize safety relevance by focusing the equivalence testing on scenarios where the AI system's performance is most critical.

Terminology

Summary

The paper addresses the gap in robustly evaluating synthetic pre-crash scenarios' practical equivalence to real-world counterparts by extending a Bayesian Region of Practical Equivalence (ROPE)-based framework with a binning-based approach.

Pruning redundant reasoning tokens

The study aims to assess whether synthetic pre-crash datasets are practically equivalent to their real-world counterparts for the intended safety impact assessment. Conventional significance testing is inadequate because it focuses on detecting differences rather than establishing practical equivalence. p. 0

Where the paper says it

However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose. Conventional significance testing is inadequate, as it focuses on detecting differences rather than establishing practical equivalence.

The research introduces two binning-based statistics to measure practically meaningful distributional differences between datasets in the context of safety impact assessment. p. 0
Where the paper says it

This study addresses the research gap by extending our previous work on a Bayesian Region of Practical Equivalence (ROPE)-based equivalence testing framework by introducing a binning-based approach to define appropriate statistics and equivalence criteria. Two binning-based statistics are proposed to measure practically meaningful distributional differences between datasets in the context of safety impact assessment.

This framework is demonstrated through a case study testing the practical equivalence of two synthetic rear-end pre-crash datasets with a previously developed reference dataset in the context of the safety impact assessment of an Automatic Emergency Braking system. p. 0
Where the paper says it

Two binning-based statistics are proposed to measure practically meaningful distributional differences between datasets in the context of safety impact assessment. The framework’s applicability is demonstrated through a case study, which tests the practical equivalence of two synthetic rear-end pre-crash datasets with a previously developed reference dataset in the context of the safety impact assessment of an Automatic Emergency Braking system.

How the training signal is built

The proposed equivalence testing method involves four steps, starting with metric selection and extraction from both datasets. Step 1 involves selecting metrics most relevant to the intended assessment to avoid false non-equivalence errors. p. 2

Where the paper says it

Step 1: Metric selection and extraction This step involves selecting the metrics most relevant to the intended assessment and extracting or deriving them from both the reference and synthetic datasets. The reason for selecting only the most relevant metrics is that, while using more metrics can make the test more comprehensive, it also heightens the risk of making a false non-equivalence error.

Step 3 requires fitting Bayesian distribution models, chosen based on expert judgment, separately to empirical distributions of the metric in reference and synthetic datasets. p. 2
Where the paper says it

, exponential, normal, log-normal, gamma, or mixture models) is chosen based on expert judgment. The models are fitted separately to empirical distributions of the metric in reference and synthetic datasets.

Posterior samples are obtained from these optimal Bayesian models fitted to the two datasets. p. 2
Where the paper says it

, 2017) is employed to select the optimal distribution model for each dataset. Posterior samples, each representing a draw of the predictive distribution parameters, are obtained from the optimal Bayesian models fitted to the two datasets.

Where the gains come from

The core of the method lies in two binning-based statistics, θ and Θ, which quantify practical differences between distributions. The statistic θ captures the maximum weighted absolute relative deviation across bins, representing a worst-case perspective. p. 3

Where the paper says it

(2) The maximum weighted absolute relative deviation across bins (i. , a worst-case perspective) is captured by θ, whereas Θ represents the total weighted absolute deviation (i.

The statistic Θ represents the total weighted absolute deviation, providing an aggregate perspective. p. 3
Where the paper says it

, a worst-case perspective) is captured by θ, whereas Θ represents the total weighted absolute deviation (i. , an aggregate perspective).

Bin weights are determined empirically by re-simulating reference scenarios with a specific DAS system to generate system-specific outcome data. p. 4
Where the paper says it

Bin Weights We propose an empirical procedure for determining bin weights: 1. Re-simulate the reference pre-crash scenarios with a virtual representation of the DAS under assessment to generate system-specific outcome data for each scenario.

The demonstration compared three datasets: the reference dataset, the PCM dataset, and an SCM-based dataset. The raw PCM dataset failed equivalence tests for Pinj and tnr because they exceeded the ROPE thresholds, reflecting a systematic bias toward more severe cases. p. 8

Where the paper says it

• Raw PCM dataset (non-equivalence): The statistics for Pinj and tnr both exceeded the ROPE thresholds, indicating systematic deviations from the reference dataset. In addition, af,min failed the equivalence test due to the large proportion of harsh-braking events in the PCM data, which shifts the distribution toward stronger deceleration values.

Conversely, the weighted PCM dataset satisfied the equivalence criteria across all four metrics, showing that careful weighting can restore distributional comparability. p. 9
Where the paper says it

• Weighted PCM dataset (equivalence): In contrast, the weighted PCM dataset satisfied the equivalence criteria across all four metrics. This finding provides evidence that carefully designed weighting strategies can restore distributional comparability between biased and representative datasets.

The SCM-based dataset failed most equivalence tests because it under-represents high-severity scenarios and over-represents cases involving a harsh-braking lead vehicle. p. 9
Where the paper says it

As shown in Fig. 7, the SCM-based dataset under-represents highseverity scenarios and over-represents cases involving a harsh-braking lead vehicle, perhaps due to the limited number of simulation runs per seed case.

The framework offers diagnostic insight by recording the weighted relative and absolute deviations of each bin, which can be used to enhance dataset generation or weighting strategies. The power analysis indicates that across all metrics and both statistics, the estimated power exceeds 0. p. 4

Where the paper says it

It is important to note that a common pair of ROPE thresholds can be applied across different metrics. This is possible because, although the bins are partitioned separately based on the reference distribution of each metric, the bin weights are consistently determined by the same weight function, ensuring comparability of the two statistics across metrics.

8 and in most cases approaches 1. p. 1
Where the paper says it

To address this gap, a promising solution lies in practical equivalence testing. It is a set of statistical approaches aimed at determining whether two treatments, processes, or groups are sufficiently similar that any observed differences are small enough to be ignored in practice (Limentani et al.

0, showing the framework reliably detects practical equivalence under adopted parameter settings. p. 4
Where the paper says it

An example of the binweight function is presented in Section 3. Parameter Settings The parameters N, α, θthd, and Θthd, described below, influence the behavior and sensitivity of the proposed testing method.

The method provides a structured workflow for practitioners to obtain interpretable, decision-relevant measures with practical equivalence between synthetic and reference data. p. 9
Where the paper says it

Beyond the specific application demonstrated here, the framework contributes methodologically by offering a structured, transparent approach to validating synthetic datasets that goes beyond conventional significance testing. By providing interpretable, relevance-weighted measures of practical difference, the framework supports both quantitative evaluation and diagnostic analysis in safety impact assessment.

The paper concludes that while appropriate weighting can mitigate bias in injury-focused datasets, it cannot compensate for all structural differences or unobserved confounders. Equivalence testing should be viewed as an iterative diagnostic tool rather than a one-off pass/fail criterion. p. 1

Where the paper says it

Although rear-end pre-crashes are used in the case study, the validation framework is inherently generic and can be readily extended to contexts beyond safety impact assessments, such as model validation in traffic flow simulation, comparative evaluation of behavioral models, or assessment of synthetic data used for system testing. By providing an interpretable and probabilistically sound foundation for assessing practical equivalence, the framework is a flexible tool for the systematic validation of synthetic data against a reference.

The framework contributes methodologically by offering a structured, transparent approach to validating synthetic datasets that goes beyond conventional significance testing. p. 9
Where the paper says it

Finally, the examples demonstrate that the proposed statistics offer diagnostic insight into which bin groups drive non-equivalence, enabling targeted identification of where synthetic datasets diverge most critically from the reference dataset. Beyond the specific application demonstrated here, the framework contributes methodologically by offering a structured, transparent approach to validating synthetic datasets that goes beyond conventional significance testing.

Future work should focus on expanding the framework to include multivariate distribution modeling to account for correlations among metrics. p. 2
Where the paper says it

These may include scenario-descriptive parameters such as initial speeds and relative positions, obtained either directly from the time series or through a parameterization of the scenario, as well as outcomebased quantities such as Delta-v or estimated injury risk (when relevant to the safety impact assessment). The framework does not depend on a specific parameterization; it accommodates any set of metrics deemed meaningful for evaluating representativeness in the intended assessment context.

The core implementation is publicly available in the bayes-binned-equivalence repository. p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

The proposed framework offers a structured workflow that practitioners can follow when validating synthetic datasets for safety impact assessment. By specifying assessment-relevant metrics and applying the binning-based statistics θ and Θ, users can obtain interpretable, decision-relevant measures with practical equivalence between synthetic and reference data. p. 1

Where the paper says it

Step 2 of the framework is intentionally flexible, allowing users to define the statistics and equivalence criteria based on the specific needs of the assessment. This study aims to provide clear and practical guidelines for implementing this step by introducing a novel approach that uses binning-based statistics and ROPEs designed in a transparent and interpretable manner (illustrated by the gray component in Fig.

The two statistics offer complementary perspectives, enabling both localized and aggregate assessment of deviations across a wide range of validation contexts. p. 3
Where the paper says it

These statistics are derived through a binning process that strikes a balance between interpretability and robustness. Interpretability is achieved by decomposing distributional differences into deviations across explicit, localized regions of the metric’s range, which can be directly related to assessment-relevant regimes.

θ highlights worst-case localized discrepancies, while Θ captures aggregate distributional differences. p. 9
Where the paper says it

To address this gap, this study proposes and demonstrates an extension of the practical equivalence testing framework for validating synthetic pre-crash scenarios to assess the safety impact of DAS, building on our earlier conference work (Wu et al. The extension introduces two binning-based statistics, θ and Θ, which quantify localized and aggregate distributional discrepancies between datasets, accounting for both the magnitude of the differences and their practical relevance.

The framework is generic and agnostic to the specific application domain because it operates on distributional representations of user-defined metrics rather than on scenario-specific assumptions. p. 2
Where the paper says it

Typically, a 95% posterior probability threshold is selected, and the ROPE is determined based on expert judgment and domain-specific knowledge relevant to the application. For a metric to be deemed practically equivalent between datasets, the 95% HDIs of both statistics’ posterior distributions must lie entirely within the defined ROPEs.

Sanity checks and plausibility checks should be performed before equivalence testing to detect implementation errors or unintended biases. Once these checks are satisfied, equivalence testing assesses representativeness relative to a reference dataset. p. 9
Where the paper says it

For (a) and (b), the top figures show the bin weights ω; the middle, the relative deviations ∆P/Pref; and the bottom, the weighted absolute relative deviations |∆P/Pref| · ω for one selected paired draw from the posterior distributions of the fitted models for the reference dataset and the dataset being compared. • Weighted PCM dataset (equivalence): In contrast, the weighted PCM dataset satisfied the equivalence criteria across all four metrics.

Sanity and plausibility checks ensure the synthetic scenarios are physically and behaviorally reasonable independent of any comparison with reference data. p. 3
Where the paper says it

The reference distribution is first partitioned into N bins, each containing (approximately) the same proportion of the total data. The same bin boundaries are then applied to the synthetic distribution, resulting in paired bin proportions that enable a direct, one-to-one comparison between the two distributions.

The framework supports the systematic refinement of synthetic datasets and their progressive alignment with assessment-relevant reference behavior. p. 9
Where the paper says it

Beyond the specific application demonstrated here, the framework contributes methodologically by offering a structured, transparent approach to validating synthetic datasets that goes beyond conventional significance testing. By providing interpretable, relevance-weighted measures of practical difference, the framework supports both quantitative evaluation and diagnostic analysis in safety impact assessment.

Table 1 summarizes the results of the equivalence tests, showing that for Pinj and tnr in the Raw PCM dataset, both statistics exceeded the ROPE thresholds, indicating systematic deviations from the reference dataset. The Weighted PCM dataset satisfied equivalence criteria across all four metrics, providing evidence that carefully designed weighting strategies can restore distributional comparability between biased and representative datasets. p. 9

Where the paper says it

• Weighted PCM dataset (equivalence): In contrast, the weighted PCM dataset satisfied the equivalence criteria across all four metrics. This finding provides evidence that carefully designed weighting strategies can restore distributional comparability between biased and representative datasets.

The SCM-based dataset failed most equivalence tests, suggesting that the apparent non-equivalence may be driven by insufficient sampling rather than model limitations. p. 9
Where the paper says it

Therefore, the present data do not permit a firm conclusion about the adequacy of the SCM; the apparent non-equivalence may be driven by insufficient sampling rather than model limitations. Finally, the examples demonstrate that the proposed statistics offer diagnostic insight into which bin groups drive non-equivalence, enabling targeted identification of where synthetic datasets diverge most critically from the reference dataset.

The framework provides diagnostic insight into which bin groups drive non-equivalence, enabling targeted identification of where synthetic datasets diverge most critically from the reference dataset. p. 9
Where the paper says it

Finally, the examples demonstrate that the proposed statistics offer diagnostic insight into which bin groups drive non-equivalence, enabling targeted identification of where synthetic datasets diverge most critically from the reference dataset. Beyond the specific application demonstrated here, the framework contributes methodologically by offering a structured, transparent approach to validating synthetic datasets that goes beyond conventional significance testing.

The code implementation is publicly available in the bayes-binned-equivalence repository. p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

The bootstrap analysis confirms that power exceeds 0. p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

8 and approaches 1. p. 1
Where the paper says it

To address this gap, a promising solution lies in practical equivalence testing. It is a set of statistical approaches aimed at determining whether two treatments, processes, or groups are sufficiently similar that any observed differences are small enough to be ignored in practice (Limentani et al.

0 across all metrics and statistics, indicating reliable detection of practical equivalence under adopted parameter settings. p. 4
Where the paper says it

Parameter Settings The parameters N, α, θthd, and Θthd, described below, influence the behavior and sensitivity of the proposed testing method. The choice of the number of bins N is particularly critical, since it directly determines the granularity of the comparison between the reference and synthetic distributions, and consequently affects the evaluation of the two statistics θ and Θ.

The paper was supported by the Fordonsstrategisk forskning och innovation (FFI) program sponsored by Vinnova. p. 0
Where the paper says it

However, real-world pre-crash data are typically limited in quantity and suffer from sampling bias and coverage issues. For example, naturalistic driving studies, such as the Second Strategic Highway Research Program (SHRP2) Naturalistic Driving Study (NDS), provide extensive exposure to everyday driving behavior, recording millions of kilometers of real-world traffic interactions (Hankey et al.

The authors thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

The references include works on statistical equivalence testing, Bayesian methods, and synthetic scenario generation for automated driving systems. p. 0
Where the paper says it

Keywords: Practical Equivalence Testing, Synthetic Pre-Crash Scenarios, Driving Automation Systems, Virtual Simulations, Safety Impact Assessment. Introduction Driving Automation Systems (DAS) (On-Road Automated Driving Committee, 2021), including Advanced Driver Assistance Systems (ADAS) and Automated Driving Systems (ADS), are expected to reduce crash risk and improve traffic safety (Pradhan et al.

The paper was submitted to Accident Analysis and Prevention on May 7, 2026. p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

p. 0
Where the paper says it

However, a critical gap remains in validating the representativeness of the resulting synthetic scenarios. In particPreprint submitted to Accident Analysis and Prevention May 7, 2026 arXiv:2605.

The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

p. 2
Where the paper says it

If each HDI lies entirely within its corresponding ROPE, the data support the practical equivalence of the metric between the compared datasets. The Python implementation of the Bayesian distribution fitting and binning-based statistics computation functions used in this study is publicly available in the bayes-binned-equivalence repository (Wu, 2026).

The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

p. 0
Where the paper says it

Practical validation of synthetic pre-crash scenarios Jian Wua,b,∗ , Ulrich Sandera , Carol Flannaganb,c, Jonas Bargman ¨ b aSafety Center, Volvo Cars, 41878 G¨oteborg, Sweden, bDepartment of Mechanics and Maritime Sciences, Chalmers University of Technology, 41756 G¨oteborg, Sweden, cUniversity of Michigan Transportation Research Institute, Ann Arbor, Michigan 48109, USA, Abstract The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios’ practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose.

The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026. The core implementation of the proposed equivalence testing procedure in this study is publicly available in the bayes-binned-equivalence repository. The authors wish to thank Mikael Ljung Aust at Volvo Cars Safety Center for reviewing the manuscript. The paper was submitted to Accident Analysis and Prevention on May 7, 2026.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Implement practical equivalence testing for synthetic pre-crash scenarios to assess whether they are similar enough for the intended assessment purpose. This allows AI systems to validate virtual simulations against real-world counterparts robustly, addressing a gap where conventional significance testing is inadequate, as it focuses on detecting differences rather than establishing practical equivalence.

  2. Utilize binning-based statistics theta and Theta to quantify practical distributional differences between datasets. This enables the AI system to move beyond simple statistical tests by providing informative quantitative assessments of practical equivalence as well as diagnostic insights into the divergence of datasets, helping identify precisely which regions of the data contribute most to non-equivalence.

  3. Incorporate bin weights derived from system-specific re-simulation outcomes, such as average re-simulated injury risk, to prioritize safety assessment relevance. This ensures that the equivalence testing focuses on conditions where the AI system is less effective at mitigating crashes, aligning validation with safety impact goals.

Abstract

The representativeness of synthetic pre-crash scenarios is crucial for assessing the safety impact of Driving Automation Systems through virtual simulations. However, a gap remains in the robust evaluation of synthetic pre-crash scenarios' practical equivalence to their real-world counterparts; that is, whether they are similar enough for the intended assessment purpose. Conventional significance testing is inadequate, as it focuses on detecting differences rather than establishing practical equivalence. This study addresses the research gap by extending our previous work on a Bayesian Region of Practical Equivalence (ROPE)-based equivalence testing framework by introducing a binning-based approach to define appropriate statistics and equivalence criteria. Two binning-based statistics are proposed to measure practically meaningful distributional differences between datasets in the context of safety impact assessment. The framework's applicability is demonstrated through a case study, which tests the practical equivalence of two synthetic rear-end pre-crash datasets with a previously developed reference dataset in the context of the safety impact assessment of an Automatic Emergency Braking system. The results show that the framework provides informative quantitative assessments of practical equivalence as well as diagnostic insights into the divergence of datasets. Although the demonstration focuses on rear-end pre-crash scenarios, the framework is generic and extensible to broader validation contexts, providing an interpretable and principled basis for practical equivalence assessment across diverse synthetic data applications.

Sources

Related papers