pith. sign in

arxiv: 2605.18823 · v1 · pith:C77QDW5Mnew · submitted 2026-05-12 · 💻 cs.LG

Multi-Pedestrian Safety Warning at Urban Intersections Use Case of Digital Twin

Yongjie Fu , Qi Gao , Mahshid Ghasemi Dehkordi , Gil Zussman , Xuan Di This is my paper

Pith reviewed 2026-05-20 22:14 UTC · model grok-4.3

classification 💻 cs.LG
keywords digital twinpedestrian safetyurban intersectionsmulti-pedestrian warningtrajectory predictionUWB sensorscamera integrationreal-time alerts
0
0 comments X

The pith

A digital twin system integrates cameras and UWB sensors to issue real-time safety warnings to pedestrians at urban intersections.

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

The paper establishes a framework for multi-pedestrian safety at urban intersections using a physical-digital twin. It combines real-world sensors with digital modeling to predict movements and send alerts via edge-cloud computing. Tests in New York City deployments and virtual reality show accurate localization, low latency, and faster user reactions to warnings. A sympathetic reader cares because this offers a practical way to enhance safety for vulnerable road users in complex traffic environments where traditional methods fall short.

Core claim

The proposed DT framework, built upon the COSMOS city-scale wireless testbed, integrates camera and ultra-wideband (UWB) technologies, edge-cloud computing, predictive trajectory modeling, and MQTT-based communication to deliver real-time safety alerts to vulnerable road users, with evaluations demonstrating high warning generation accuracy, localization accuracy, efficient end-to-end latency, and significant reductions in user response time.

What carries the argument

The tightly coupled physical-digital twin framework that fuses camera and UWB data for trajectory prediction and warning delivery.

If this is right

  • The system achieves high warning generation accuracy in real conditions.
  • Localization accuracy is sufficient for timely interventions.
  • Efficient end-to-end latency supports different model configurations.
  • User response time decreases significantly when warnings are issued.
  • The framework is scalable, modular, and generalizable for other urban safety applications.

Where Pith is reading between the lines

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

  • Extending the framework to integrate with vehicle sensors could create comprehensive intersection management.
  • Deployment across multiple intersections might enable city-scale safety networks using existing testbeds.
  • Further validation in adverse weather or high-density scenarios would test the limits of the sensor fusion.
  • Potential for combining with machine learning models to improve prediction accuracy over time.

Load-bearing premise

The tightly coupled physical-digital twin can reliably predict trajectories and issue timely warnings under real urban conditions without major sensor interference or modeling errors.

What would settle it

An experiment showing that in a crowded intersection, the system issues false negatives or warnings too late due to UWB signal interference or camera occlusion, leading to no reduction in response time.

Figures

Figures reproduced from arXiv: 2605.18823 by Gil Zussman, Mahshid Ghasemi Dehkordi, Qi Gao, Xuan Di, Yongjie Fu.

Figure 1
Figure 1. Figure 1: Overview of the proposed digital twin pipeline and its intersection safety warning application. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Traffic statistics for NYC’s intersection of 120th St. & [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Demonstration of UWB localization We employ UWB technology for pedestrian localization using the native Two-Way Ranging (TWR) protocol. TWR estimates distance from the round-trip signal time between anchors and users without requiring clock synchronization. Under line-of-sight conditions, it achieves sub-10 cm accuracy in a plug-and-play configuration, making it suitable for on edge localization [PITH_FUL… view at source ↗
Figure 4
Figure 4. Figure 4: A simulated pedestrian in CARLA (left) surrounded [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: ROC curves for trajectory prediction to determine [PITH_FULL_IMAGE:figures/full_fig_p004_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Performance comparison for trajectory prediction [PITH_FULL_IMAGE:figures/full_fig_p005_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Response time determination in the VR experiment [PITH_FULL_IMAGE:figures/full_fig_p005_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Survey results from the VR experiment categorize [PITH_FULL_IMAGE:figures/full_fig_p006_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Comparison of response times with and without a [PITH_FULL_IMAGE:figures/full_fig_p006_9.png] view at source ↗
read the original abstract

Digital twins (DTs) for urban transportation systems have gained increasing attention; however, their systematic evaluation in safety-critical scenarios remains limited. This paper presents a multi-pedestrian safety warning system at urban intersections enabled by a tightly coupled physical-digital twin framework. Built upon the COSMOS city-scale wireless testbed in New York City, the proposed system integrates camera and ultra-wideband (UWB), edge-cloud computing, predictive trajectory modeling, and MQTT-based communication to deliver real-time safety alerts to vulnerable road users (VRUs). The system is evaluated through both field deployment and virtual reality (VR) experiments. Results demonstrate high warning generation accuracy, localization accuracy, efficient end-to-end latency under different model configurations, and significant reductions in user response time when warnings are issued. The proposed DT framework provides a scalable, modular, and generalizable solution for real-time multi-pedestrian safety enhancement at complex urban intersections.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 1 minor

Summary. The paper presents a multi-pedestrian safety warning system at urban intersections enabled by a tightly coupled physical-digital twin framework. Built on the COSMOS city-scale wireless testbed in New York City, it integrates camera and ultra-wideband (UWB) sensors, edge-cloud computing, predictive trajectory modeling, and MQTT-based communication. The system is evaluated through field deployment and virtual reality experiments, with results claiming high warning generation accuracy, localization accuracy, efficient latency under different configurations, and reductions in user response time. The authors conclude that the DT framework provides a scalable, modular, and generalizable solution for real-time safety enhancement.

Significance. If the quantitative performance claims can be substantiated, this work offers a concrete example of applying digital twins to enhance pedestrian safety in complex urban environments using existing testbed infrastructure. The integration of multiple sensing modalities and real-time communication is noteworthy. However, the absence of detailed metrics and cross-validation experiments weakens the generalizability assertion, which is central to the paper's positioning.

major comments (2)
  1. [Abstract] The claims of 'high warning generation accuracy', 'localization accuracy', and 'efficient end-to-end latency' are stated without any numerical values, standard deviations, or comparisons to baselines or alternative methods, making it challenging to evaluate the effectiveness of the proposed system.
  2. [Results and Evaluation] The assertion that the framework is 'scalable, modular, and generalizable' is not supported by evidence beyond the single COSMOS NYC testbed deployment; no experiments on different urban intersections, varying weather conditions, or sensor densities are reported to demonstrate transferability.
minor comments (1)
  1. [Methodology] The description of the predictive trajectory modeling could benefit from more details on the specific algorithms used and any assumptions made about pedestrian behavior.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments and the opportunity to improve the manuscript. We address each major comment below and indicate the revisions we will make.

read point-by-point responses

  1. Referee: [Abstract] The claims of 'high warning generation accuracy', 'localization accuracy', and 'efficient end-to-end latency' are stated without any numerical values, standard deviations, or comparisons to baselines or alternative methods, making it challenging to evaluate the effectiveness of the proposed system.

    Authors: We agree that the abstract would benefit from greater specificity. The results section of the manuscript reports concrete metrics for warning generation accuracy, localization error, and end-to-end latency under multiple configurations, including comparisons to non-predictive baselines. We will revise the abstract to include representative numerical values and standard deviations drawn directly from those experiments. revision: yes

  2. Referee: [Results and Evaluation] The assertion that the framework is 'scalable, modular, and generalizable' is not supported by evidence beyond the single COSMOS NYC testbed deployment; no experiments on different urban intersections, varying weather conditions, or sensor densities are reported to demonstrate transferability.

    Authors: We acknowledge that the empirical evaluation is confined to the COSMOS NYC deployment and that additional cross-site or cross-condition experiments would strengthen claims of transferability. The manuscript demonstrates modularity through the explicit separation of sensing, prediction, edge-cloud, and communication modules. We will revise the discussion section to qualify the generalizability statement, highlight the architectural features that enable adaptation to other environments, and explicitly note the current evaluation scope as a limitation with directions for future work. revision: partial

Circularity Check

0 steps flagged

No derivation chain or fitted predictions; implementation and empirical evaluation only

full rationale

The paper presents a deployed system on the COSMOS testbed integrating cameras, UWB, edge-cloud computing, trajectory modeling, and MQTT for real-time warnings, evaluated via field tests and VR experiments. No equations, parameter fitting, or predictive derivations appear in the provided abstract or description. Claims of scalability and generalizability rest on the modular architecture and single-site results rather than any self-referential reduction, self-citation load-bearing argument, or renaming of prior results. This matches the reader's assessment of an implemented system without circular modeling steps.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

This is an applied engineering paper focused on system integration and experimental evaluation. No mathematical derivations, free parameters, axioms, or new postulated entities appear in the abstract.

pith-pipeline@v0.9.0 · 5700 in / 1045 out tokens · 38004 ms · 2026-05-20T22:14:34.333254+00:00 · methodology

discussion (0)

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

Lean theorems connected to this paper

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

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

Reference graph

Works this paper leans on

18 extracted references · 18 canonical work pages

  1. [1]

    Vulnerable road user safety assessment guidance,

    USDOT FHW A, “Vulnerable road user safety assessment guidance,” Memorandum, 2022

    work page 2022

  2. [2]

    Real-time video analytics for urban safety: Deployment over edge and end devices,

    M. Ghasemi, Y . Fu, X. Ouyang, P. Wang, M. Turkcan, J. Tavori, S. Kleis- archaki, T. Calmant, L. G ¨urgen, Z. Kostic, X. Di, G. Zussman, and J. Ghaderi, “Real-time video analytics for urban safety: Deployment over edge and end devices,” inProceedings of the ACM/IEEE Symposium on Edge Computing (SEC), 2025

    work page 2025

  3. [3]

    Trajectron++: Dynamically-feasible trajectory forecasting with heterogeneous data,

    T. Salzmann, B. Ivanovic, P. Chakravarty, and M. Pavone, “Trajectron++: Dynamically-feasible trajectory forecasting with heterogeneous data,” in European Conference on Computer Vision (ECCV), 2020

    work page 2020

  4. [4]

    Smart mobility digital twin based automated vehicle navigation system: A proof of concept,

    K. Wang, Z. Li, K. Nonomura, T. Yu, K. Sakaguchi, O. Hashash, and W. Saad, “Smart mobility digital twin based automated vehicle navigation system: A proof of concept,”IEEE Transactions on Intelligent V ehicles, 2024

    work page 2024

  5. [5]

    A novel cooperative localization method based on imu and uwb,

    Y . Han, C. Wei, R. Li, J. Wang, and H. Yu, “A novel cooperative localization method based on imu and uwb,”Sensors, vol. 20, no. 2, p. 467, 2020

    work page 2020

  6. [6]

    Design of the risk warning system based on uwb for the positioning of substation staff,

    Y . Yu, Z. Li, Z. Wang, and C. Li, “Design of the risk warning system based on uwb for the positioning of substation staff,” in2021 8th International F orum on Electrical Engineering and Automation (IFEEA), 2021, pp. 148–153

    work page 2021

  7. [7]

    Enhancing safety on construction sites: A uwb-based prox- imity warning system ensuring gdpr compliance to prevent collision hazards,

    S. M. Ventura, P. Bellagente, S. Rinaldi, A. Flammini, and A. L. C. Ciribini, “Enhancing safety on construction sites: A uwb-based prox- imity warning system ensuring gdpr compliance to prevent collision hazards,”Sensors, vol. 23, no. 24, p. 9770, 2023

    work page 2023

  8. [8]

    Digital twin for pedestrian safety warning at a single urban traffic intersection,

    Y . Fu, M. K. Turkcan, V . Anantha, Z. Kostic, G. Zussman, and X. Di, “Digital twin for pedestrian safety warning at a single urban traffic intersection,” in2024 IEEE Intelligent V ehicles Symposium (IV). IEEE, 2024, pp. 2640–2645

    work page 2024

  9. [9]

    Ai-powered cps-enabled vulnerable- user-aware urban transportation digital twin: Methods and applications,

    Y . Fu, M. K. Turkcan, M. Ghasemi, Z. Mo, C. Zang, A. Adhikari, Z. Kostic, G. Zussman, and X. Di, “Ai-powered cps-enabled vulnerable- user-aware urban transportation digital twin: Methods and applications,” IEEE Transactions on Intelligent Transportation Systems, 2026

    work page 2026

  10. [10]

    Challenge: Cosmos: A city-scale programmable testbed for experimentation with advanced wireless,

    D. Raychaudhuri, I. Seskar, G. Zussman, T. Korakis, D. Kilper, T. Chen, J. Kolodziejski, M. Sherman, Z. Kostic, X. Guet al., “Challenge: Cosmos: A city-scale programmable testbed for experimentation with advanced wireless,” inProceedings of the 26th annual international conference on mobile computing and networking, 2020, pp. 1–13

    work page 2020

  11. [11]

    FHW A, https://highways.dot.gov/safety/intersection-safety/about#: ∼: text=Intersecting%20roadways%20are%20necessary%20to,program% 20focus%20area%20for%20FHW A., 2024

    work page 2024

  12. [12]

    Infostgcan: An information-maximizing spatial- temporal graph convolutional attention network for heterogeneous hu- man trajectory prediction,

    K. Ruan and X. Di, “Infostgcan: An information-maximizing spatial- temporal graph convolutional attention network for heterogeneous hu- man trajectory prediction,”Computers, vol. 13, no. 6, p. 151, 2024

    work page 2024

  13. [13]

    Pi-neugode: Physics-informed graph neural ordinary differential equations for spatiotemporal trajectory prediction,

    Z. Mo, Y . Fu, and X. Di, “Pi-neugode: Physics-informed graph neural ordinary differential equations for spatiotemporal trajectory prediction,” inProceedings of the 23rd International Conference on Autonomous Agents and Multiagent Systems, 2024, pp. 1418–1426

    work page 2024

  14. [14]

    Uncertainty quantification of car-following behaviors: physics-informed generative adversarial networks,

    Z. Mo and X. Di, “Uncertainty quantification of car-following behaviors: physics-informed generative adversarial networks,” inthe 28th ACM SIGKDD in conjunction with the 11th International Workshop on Urban Computing (UrbComp2022), 2022

    work page 2022

  15. [15]

    Trajectron++: Dynamically-feasible trajectory forecasting with heterogeneous data,

    T. Salzmann, B. Ivanovic, P. Chakravarty, and M. Pavone, “Trajectron++: Dynamically-feasible trajectory forecasting with heterogeneous data,” in Computer Vision–ECCV 2020. Springer, 2020, pp. 683–700

    work page 2020

  16. [16]

    YOLO by Ultralytics,

    G. Jocher, A. Chaurasia, and J. Qiu, “YOLO by Ultralytics,” 2023. [Online]. Available: https://github.com/ultralytics/ultralytics

    work page 2023

  17. [17]

    Modeling the impact of wi-fi 6e traffic on ultra-wideband communication performance,

    M. Stocker, A. B ¨ogl, E. Ember, M. Schuh, K. R ¨omer, C. A. Boano, S. Tertinek, and P. C. Pelegr ´ın, “Modeling the impact of wi-fi 6e traffic on ultra-wideband communication performance,” in2025 International Conference on Modeling, Analysis and Simulation of Wireless and Mobile Systems (MSWiM). IEEE, 2025, pp. 54–63

    work page 2025

  18. [18]

    Unity (version 2023),

    Unity Technologies, “Unity (version 2023),” https://unity.com/, 2024

    work page 2023