pith. sign in

arxiv: 1907.07495 · v1 · pith:JSPKTFAOnew · submitted 2019-07-17 · 📡 eess.SP · math.OC

Combining traffic counts and Bluetooth data for link-origin-destination matrix estimation in large urban networks: The Brisbane case study

Gabriel Michau , Nelly Pustelnik , Pierre Borgnat , Patrice Abry , Ashish Bhaskar , Edward Chung This is my paper

Pith reviewed 2026-05-24 20:08 UTC · model grok-4.3

classification 📡 eess.SP math.OC
keywords origin-destination matrixtraffic estimationBluetooth datainverse problemurban traffic networkslink-origin-destination matrixBrisbane case study
0
0 comments X

The pith

Bluetooth data paired with traffic counts enables estimation of three-dimensional link-origin-destination matrices.

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

The paper establishes a method to estimate link-origin-destination matrices by fusing traffic counts with Bluetooth detections that reveal partial origins, destinations and routes for some vehicles. Traditional origin-destination matrices are extended into three dimensions so that traffic assignment on specific links is also recovered. The estimation is cast as an inverse problem whose objective function balances multiple traffic properties such as conservation and consistency with observed counts. A large-scale case study on Brisbane's network of over 600 Bluetooth detectors demonstrates how the resulting matrices support detailed traffic analysis. A sympathetic reader would care because such matrices offer a more complete picture of urban movement than counts or surveys alone can provide.

Core claim

Link-origin-destination matrices can be recovered by solving an inverse problem that incorporates both aggregate traffic counts and the partial origin-destination and itinerary information supplied by Bluetooth detectors; the objective function encodes a trade-off among traffic conservation, consistency with measurements and other structural properties, and the approach is implemented and illustrated on the Brisbane urban network.

What carries the argument

The inverse-problem formulation whose objective function represents a trade-off between important properties the traffic has to satisfy, producing the three-dimensional link-origin-destination matrix.

If this is right

  • The resulting matrices supply link-level assignment information in addition to origin-destination totals.
  • The method scales to networks instrumented with hundreds of Bluetooth detectors.
  • Traffic analysis gains new granularity for planning and operations once the matrices are obtained.
  • The same inverse-problem structure can accept other partial itinerary data sources.

Where Pith is reading between the lines

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

  • If the weighting parameters can be learned from limited ground-truth data, the approach might generalize across cities with different detector densities.
  • The three-dimensional matrices could feed directly into dynamic traffic assignment models or real-time control algorithms.
  • Comparison of successive daily matrices might reveal recurring route-choice patterns without additional modeling.

Load-bearing premise

The chosen weighting of the objective-function terms and the exact definitions of the traffic properties yield a matrix that is both valid and useful without further external checks.

What would settle it

Independent traffic observations or surveys in Brisbane that show systematic mismatch between the estimated link flows or origin-destination volumes and the held-out measurements.

Figures

Figures reproduced from arXiv: 1907.07495 by Ashish Bhaskar, Edward Chung, Gabriel Michau, Nelly Pustelnik, Patrice Abry, Pierre Borgnat.

Figure 1
Figure 1. Figure 1: Traffic estimation: Traditional framework (a) versus proposed one (b). In blue boxes are inputs; [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: (a) Traffic counts derived from induction loop detectors after transfer to the simplified network. [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Distribution of traffic flows in the LODM, (a) for [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: (a) Brisbane Traffic Counts on the simplified network for the area of study. (b) Origin-Destination [PITH_FULL_IMAGE:figures/full_fig_p013_4.png] view at source ↗
read the original abstract

Origin-Destination matrix estimation is a keystone for traffic representation and analysis. Traditionally estimated thanks to traffic counts, surveys and socio-economic models, recent technological advances permit to rethink the estimation problem. Road user identification technologies, such as connected GPS, Bluetooth or Wifi detectors bring additional information, that is, for a fraction of the users, the origin, the destination and to some extend the itinerary taken. In the present work, this additional information is used for the estimation of a more comprehensive traffic representation tool: the link-origin-destination matrix. Such three-dimensional matrices extend the concept of traditional origin-destination matrices by also giving information on the traffic assignment. Their estimation is solved as an inverse problem whose objective function represents a trade-off between important properties the traffic has to satisfy. This article presents the theory and how to implement such method on real dataset. With the case study of Brisbane City where over 600 hundreds Bluetooth detectors have been installed it also illustrates the opportunities such link-origin-destination matrices create for traffic analysis.

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 claims to develop and implement a method for estimating three-dimensional link-origin-destination (link-OD) matrices by combining traffic counts with Bluetooth-derived origin, destination, and partial path information. The estimation is formulated as an inverse problem whose objective function trades off consistency with observed counts, Bluetooth paths, and other traffic properties; the approach is demonstrated on the Brisbane network using data from over 600 Bluetooth detectors, with discussion of resulting opportunities for traffic analysis.

Significance. If the estimated link-OD matrices can be shown to recover actual flows more accurately than count-only baselines, the work would provide a practical way to obtain assignment-aware traffic representations at urban scale. The deployment of a large real-world Bluetooth sensor network and the explicit treatment of implementation details on that dataset constitute a concrete strength.

major comments (2)
  1. [Case study and results (likely §4–5)] The central construction solves the inverse problem by minimizing an objective that trades off consistency with traffic counts, Bluetooth-derived paths, and other traffic properties, yet the manuscript supplies no quantitative comparison of the resulting link-OD entries to independent survey or probe-vehicle OD data, nor any cross-validation that would show the chosen weights recover known matrices better than count-only baselines.
  2. [Method (objective-function formulation)] The objective-function weights and constraint definitions are presented as producing a valid matrix, but no external validation or sensitivity analysis against ground-truth matrices is reported; this is load-bearing for the claim that the estimated matrices constitute a useful traffic representation tool.
minor comments (1)
  1. [Abstract] Abstract contains the phrase 'over 600 hundreds Bluetooth detectors', which is a typographical error.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the positive assessment of the real-world deployment and implementation details. We address the two major comments below, which both concern validation.

read point-by-point responses

  1. Referee: [Case study and results (likely §4–5)] The central construction solves the inverse problem by minimizing an objective that trades off consistency with traffic counts, Bluetooth-derived paths, and other traffic properties, yet the manuscript supplies no quantitative comparison of the resulting link-OD entries to independent survey or probe-vehicle OD data, nor any cross-validation that would show the chosen weights recover known matrices better than count-only baselines.

    Authors: We agree that a direct quantitative comparison to independent ground-truth OD data would strengthen the results. No such comprehensive link-level survey or probe-vehicle dataset exists for the Brisbane network at the required scale and granularity. The presented validation relies on internal consistency with the input counts and Bluetooth paths. In revision we will add an explicit discussion of this limitation together with a sensitivity study on the objective weights. revision: partial

  2. Referee: [Method (objective-function formulation)] The objective-function weights and constraint definitions are presented as producing a valid matrix, but no external validation or sensitivity analysis against ground-truth matrices is reported; this is load-bearing for the claim that the estimated matrices constitute a useful traffic representation tool.

    Authors: The formulation guarantees feasibility with respect to the observed counts and Bluetooth information by construction. We acknowledge that external validation against held-out ground-truth matrices is absent. We will incorporate a sensitivity analysis on the weighting parameters in the revised manuscript to address robustness. revision: yes

standing simulated objections not resolved
  • No independent ground-truth link-OD survey or probe data is available for the Brisbane network, preventing quantitative comparison or cross-validation against count-only baselines.

Circularity Check

0 steps flagged

No significant circularity; estimation reduces to standard inverse-problem optimization on external data inputs.

full rationale

The paper formulates link-OD matrix estimation as minimization of an objective trading off consistency with observed traffic counts and Bluetooth-derived paths. No quoted step shows a fitted parameter renamed as a prediction, a self-citation chain supplying the uniqueness or ansatz, or any result defined in terms of itself. The derivation remains self-contained against the supplied count and detector data; external validation is a separate correctness issue, not circularity.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim rests on the assumption that traffic obeys a set of properties that can be encoded as an objective function whose weights produce a meaningful matrix; no new physical entities are introduced.

free parameters (1)
  • objective-function weights
    Trade-off parameters balancing the listed traffic properties are introduced to define the inverse problem.
axioms (1)
  • domain assumption Traffic must satisfy a collection of properties (flow conservation, consistency with counts, etc.) that can be expressed in a single scalar objective.
    The inverse-problem formulation in the abstract is defined by this trade-off.

pith-pipeline@v0.9.0 · 5730 in / 1169 out tokens · 18960 ms · 2026-05-24T20:08:02.199860+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

44 extracted references · 44 canonical work pages

  1. [1]

    Methods for the study of retail relationships

    W. J. Reilly et al., “Methods for the study of retail relationships”, 1929

    work page 1929

  2. [2]

    What’s missing from australian household travel surveys? Off-peak travel!

    T. Veitch, M. Paech, and J. Eaton, “What’s missing from australian household travel surveys? Off-peak travel!”, in Australasian Transport Research Forum 2013, Brisbane, Oct. 2013

    work page 2013

  3. [3]

    Towards a generic benchmarking platform for origin–destination flows es- timation/updating algorithms: Design, demonstration and validation

    C. Antoniou, J. Barcel´ o, M. Breen, M. Bullejos, J. Casas, E. Cipriani, B. Ciuffo, T. Djukic, S. Hoogen- doorn, V. Marzano, et al., “Towards a generic benchmarking platform for origin–destination flows es- timation/updating algorithms: Design, demonstration and validation”, Transportation Research Part C: Emerging Technologies, vol. 66, pp. 79–98, 2016

    work page 2016

  4. [4]

    On combining maximum entropy trip matrix estimation with user optimal assignment

    C. Fisk, “On combining maximum entropy trip matrix estimation with user optimal assignment”, Transportation Research Part B: Methodological, vol. 22, no. 1, pp. 69–73, 1988

    work page 1988

  5. [5]

    Entropy in urban and regional modelling: Retrospect and prospect

    A. Wilson, “Entropy in urban and regional modelling: Retrospect and prospect.”, Geographical Anal- ysis, vol. 42, no. 4, pp. 364–394, 2010

    work page 2010

  6. [6]

    The most likely trip matrix estimated from traffic counts

    H. J. Van Zuylen and L. G. Willumsen, “The most likely trip matrix estimated from traffic counts”, Transportation Research Part B: Methodological, vol. 14, no. 3, pp. 281–293, 1980

    work page 1980

  7. [7]

    Estimation of origin-destination matrix from traffic counts: A comparison of entropy maximizing and information minimizing models

    W. H. Lam and H. Lo, “Estimation of origin-destination matrix from traffic counts: A comparison of entropy maximizing and information minimizing models”, Transportation Planning and Technology , vol. 16, no. 2, pp. 85–104, 1991

    work page 1991

  8. [8]

    A maximum likelihood model for estimating origin-destination matrices

    H. Spiess, “A maximum likelihood model for estimating origin-destination matrices”, Transportation Research Part B: Methodological, vol. 21, no. 5, pp. 395–412, 1987

    work page 1987

  9. [9]

    Methods to combine different data sources and estimate origin-destination matrices

    M. Ben-Akiva, “Methods to combine different data sources and estimate origin-destination matrices”, Transportation and traffic theory, 1987

    work page 1987

  10. [10]

    A bayesian method for estimating traffic flows based on plate scanning

    E. Castillo, P. Jim´ enez, J. M. Men´ endez, and M. Nogal, “A bayesian method for estimating traffic flows based on plate scanning”, Transportation, vol. 40, no. 1, pp. 173–201, 2013

    work page 2013

  11. [11]

    Bayesian inference for transportation origin– destination matrices: The poisson–inverse gaussian and other poisson mixtures

    K. Perrakis, D. Karlis, M. Cools, and D. Janssens, “Bayesian inference for transportation origin– destination matrices: The poisson–inverse gaussian and other poisson mixtures”, Journal of the Royal Statistical Society: Series A (Statistics in Society) , vol. 178, no. 1, pp. 271–296, 2015

    work page 2015

  12. [12]

    Inferences on trip matrices from observations on link volumes: A bayesian statistical approach

    M. Maher, “Inferences on trip matrices from observations on link volumes: A bayesian statistical approach”, Transportation Research Part B: Methodological, vol. 17, no. 6, pp. 435–447, 1983

    work page 1983

  13. [13]

    Estimation of origin–destination matrices from link counts and sporadic routing data

    K. Parry and M. L. Hazelton, “Estimation of origin–destination matrices from link counts and sporadic routing data”, Transportation Research Part B: Methodological, vol. 46, no. 1, pp. 175–188, 2012

    work page 2012

  14. [14]

    An iterative assignment approach to capacity restraint on arterial networks

    R. Smock, “An iterative assignment approach to capacity restraint on arterial networks”, Highway Research Board Bulletin, no. 347, 1962

    work page 1962

  15. [15]

    Estimation of an od matrix from traffic counts–a review

    L. G. Willumsen, “Estimation of an od matrix from traffic counts–a review.”, 1978

    work page 1978

  16. [16]

    Iterative update of route choice proportions in od estimation

    M. Yousefikia, A. R. Mamdoohi, and M. Noruzoliaee, “Iterative update of route choice proportions in od estimation”, in Proceedings of the Institution of Civil Engineers-Transport , Thomas Telford Ltd, vol. 169, 2016, pp. 53–60

    work page 2016

  17. [17]

    Estimation of origin-destination matrices from link traffic counts on congested networks

    H. Yang, T. Sasaki, Y. Iida, and Y. Asakura, “Estimation of origin-destination matrices from link traffic counts on congested networks”, Transportation Research Part B: Methodological, vol. 26, no. 6, pp. 417–434, 1992

    work page 1992

  18. [18]

    Estimation of trip matrices from traffic counts and survey data: A generalized least squares estimator

    E. Cascetta, “Estimation of trip matrices from traffic counts and survey data: A generalized least squares estimator”, Transportation Research Part B: Methodological , vol. 18, no. 4-5, pp. 289–299, 1984

    work page 1984

  19. [19]

    A modified physarum-inspired model for the user equilibrium traffic assignment problem

    S. Xu, W. Jiang, X. Deng, and Y. Shou, “A modified physarum-inspired model for the user equilibrium traffic assignment problem”, Applied Mathematical Modelling, vol. 55, pp. 340–353, 2018

    work page 2018

  20. [20]

    A linear programming approach for synthesizing origin-destination trip tables from link traffic volumes

    H. D. Sherali, R. Sivanandan, and A. G. Hobeika, “A linear programming approach for synthesizing origin-destination trip tables from link traffic volumes”, Transportation Research Part B: Methodolog- ical, vol. 28, no. 3, pp. 213–233, 1994

    work page 1994

  21. [21]

    Simultaneous estimation of the origin-destination matrices and travel-cost coefficient for congested networks in a stochastic user equilibrium

    H. Yang, Q. Meng, and M. G. Bell, “Simultaneous estimation of the origin-destination matrices and travel-cost coefficient for congested networks in a stochastic user equilibrium”, Transportation Science, vol. 35, no. 2, pp. 107–123, 2001

    work page 2001

  22. [22]

    A bi-level programming approach for trip matrix estimation and traffic control problems with stochastic user equilibrium link flows

    M. J. Maher, X. Zhang, and D. Van Vliet, “A bi-level programming approach for trip matrix estimation and traffic control problems with stochastic user equilibrium link flows”, Transportation Research Part B: Methodological, vol. 35, no. 1, pp. 23–40, 2001. 14 A preprint - May 31, 2021

    work page 2001

  23. [23]

    Adjustment of o–d trip matrices from observed volumes: An algorithmic approach based on conjugate directions

    E. Codina and J. Barcel´ o, “Adjustment of o–d trip matrices from observed volumes: An algorithmic approach based on conjugate directions”, European Journal of Operational Research , vol. 155, no. 3, pp. 535–557, 2004

    work page 2004

  24. [24]

    On the calibration of the combined distribution-assignment model

    S. Erlander, S. Nguyen, and N. F. Stewart, “On the calibration of the combined distribution-assignment model”, Transportation Research Part B: Methodological, vol. 13, no. 3, pp. 259–267, 1979

    work page 1979

  25. [25]

    Network model calibration studies

    T. Toledo, T. Kolechkina, P. Wagner, B. Ciuffo, C. Azevedo, V. Marzano, and G. Fl¨ otter¨ od, “Network model calibration studies”, Daamen W., Buisson C. and S. Hoogendoorn (eds). Traffic simulation and data: Validation methods and applications, CRC Press, Taylor and Francis, London, pp. 141–162, 2015

    work page 2015

  26. [26]

    A robust framework for the estimation of dynamic od trip matrices for reliable traffic management

    J. Barcel´ o and L. Montero, “A robust framework for the estimation of dynamic od trip matrices for reliable traffic management”, Transportation Research Procedia, vol. 10, pp. 134–144, 2015

    work page 2015

  27. [27]

    Quasi-dynamic estimation of od flows from traffic counts without prior od matrix

    D. Bauer, G. Richter, J. Asamer, B. Heilmann, G. Lenz, and R. K¨ olbl, “Quasi-dynamic estimation of od flows from traffic counts without prior od matrix”, IEEE Transactions on Intelligent Transportation Systems, vol. 19, no. 6, pp. 2025–2034, 2017

    work page 2025

  28. [28]

    Updating origin–destination matrices with aggregated data of gps traces

    Q. Ge and D. Fukuda, “Updating origin–destination matrices with aggregated data of gps traces”, Transportation Research Part C: Emerging Technologies, vol. 69, pp. 291–312, 2016

    work page 2016

  29. [29]

    Origin-destination estimation using probe vehicle trajectory and link counts

    X. Yang, Y. Lu, and W. Hao, “Origin-destination estimation using probe vehicle trajectory and link counts”, Journal of Advanced Transportation, vol. 2017, 2017

    work page 2017

  30. [30]

    Optimal traffic sensor location for origin–destination estimation using a compressed sensing framework

    P. Ye and D. Wen, “Optimal traffic sensor location for origin–destination estimation using a compressed sensing framework”, IEEE Transactions on Intelligent Transportation Systems, vol. 18, no. 7, pp. 1857– 1866, 2016

    work page 2016

  31. [31]

    A primal-dual algorithm for link dependent origin destination matrix estimation

    G. Michau, N. Pustelnik, P. Borgnat, P. Abry, A. Nantes, A. Bhaskar, and E. Chung, “A primal-dual algorithm for link dependent origin destination matrix estimation”, IEEE Transactions on Signal and Information Processing over Networks , vol. 3, no. 1, pp. 104–113, 2016

    work page 2016

  32. [32]

    Fundamental understanding on the use of bluetooth scanner as a comple- mentary transport data

    A. Bhaskar and E. Chung, “Fundamental understanding on the use of bluetooth scanner as a comple- mentary transport data”, Transportation Research Part C: Emerging Technologies, vol. 37, pp. 42–72, 2013

    work page 2013

  33. [33]

    Bluetooth data in an urban context: Retrieving vehicle trajectories

    G. Michau, A. Nantes, A. Bhaskar, E. Chung, P. Abry, and P. Borgnat, “Bluetooth data in an urban context: Retrieving vehicle trajectories”, IEEE Transactions on Intelligent Transportation Systems , vol. 18, no. 9, pp. 2377–2386, 2017

    work page 2017

  34. [34]

    Evaluation of traffic data obtained via gps-enabled mobile phones: The mobile century field experiment

    J. C. Herrera, D. B. Work, R. Herring, X. J. Ban, Q. Jacobson, and A. M. Bayen, “Evaluation of traffic data obtained via gps-enabled mobile phones: The mobile century field experiment”, Transportation Research Part C: Emerging Technologies, vol. 18, no. 4, pp. 568–583, 2010

    work page 2010

  35. [35]

    Estimating route choice and travel time reliability with field observations of bluetooth probe vehicles

    A. M. Hainen, J. S. Wasson, S. M. Hubbard, S. M. Remias, G. D. Farnsworth, and D. M. Bullock, “Estimating route choice and travel time reliability with field observations of bluetooth probe vehicles”, Transportation Research Record, vol. 2256, no. 1, pp. 43–50, 2011

    work page 2011

  36. [36]

    Vehicle trajectory reconstruction using automatic vehicle identification and traffic count data

    Y. Feng, J. Sun, and P. Chen, “Vehicle trajectory reconstruction using automatic vehicle identification and traffic count data”, Journal of advanced transportation, vol. 49, no. 2, pp. 174–194, 2015

    work page 2015

  37. [37]

    Evaluation of trade-offs between two data sources for the accurate estimation of origin–destination matrices

    P. G´ omez, M. Men´ endez, and E. M´ erida-Casermeiro, “Evaluation of trade-offs between two data sources for the accurate estimation of origin–destination matrices”, Transportmetrica B: Transport Dynamics, vol. 3, no. 3, pp. 222–245, 2015

    work page 2015

  38. [38]

    Retrieving trip information from a discrete detectors network: The case of brisbane bluetooth detectors

    G. Michau, A. Nantes, E. Chung, P. Abry, and P. Borgnat, “Retrieving trip information from a discrete detectors network: The case of brisbane bluetooth detectors”, in Proceedings of CAITR, Sydney, 02 Feb. 2014, Sydney, 2014

    work page 2014

  39. [39]

    Generating route-specific origin–destination tables using bluetooth technology

    C. Carpenter, M. Fowler, and T. J. Adler, “Generating route-specific origin–destination tables using bluetooth technology”, Transportation Research Record, vol. 2308, no. 1, pp. 96–102, 2012

    work page 2012

  40. [40]

    Understanding road usage patterns in urban areas

    P. Wang, T. Hunter, A. M. Bayen, K. Schechtner, and M. C. Gonz´ alez, “Understanding road usage patterns in urban areas”, Scientific reports, vol. 2, p. 1001, 2012

    work page 2012

  41. [41]

    Estimation of origin-destination matrix from traffic counts: The state of the art

    S. Bera and K. Rao, “Estimation of origin-destination matrix from traffic counts: The state of the art”, 2011

    work page 2011

  42. [42]

    Nonsmooth convex opti- mization for structured illumination microscopy image reconstruction

    J. Boulanger, N. Pustelnik, L. Condat, L. Sengmanivong, and T. Piolot, “Nonsmooth convex opti- mization for structured illumination microscopy image reconstruction”, Inverse problems, vol. 34, no. 9, p. 095 004, 2018

    work page 2018

  43. [43]

    Le Gall, Brownian motion, martingales, and stochastic calculus

    J.-F. Le Gall, Brownian motion, martingales, and stochastic calculus . Springer, 2016, vol. 274

    work page 2016

  44. [44]

    A douglas-rachford splitting approach to nonsmooth convex variational signal recovery

    P. L. Combettes and J.-C. Pesquet, “A douglas-rachford splitting approach to nonsmooth convex variational signal recovery”, IEEE Journal of Selected Topics in Signal Processing, vol. 1, no. 4, p. 564, 2007. 15

    work page 2007