{"id":"00cee1e1-dd5b-44ff-ae3c-93d9e18faca5","arxiv_id":"2505.15613","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Simulations on real urban street networks reveal congestion phase transitions similar to a square grid, an extra semi-congested regime in organically grown cities, and higher extreme event probability at low-degree nodes.","lead":"This paper simulates vehicle traffic on real street networks of four cities and on a square grid, using random-walk particles with queues and birth-death dynamics. It finds that organically grown cities show a mixed semi-congested state that planned grids do not, and that smaller intersections experience extreme congestion spikes more often than major hubs.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed semi-congested regime and phase boundaries rest on single finite-time order-parameter estimates with no error bars or finite-size scaling; a slowly congesting transient could masquerade as 0<rho<1.","rationale":"The paper's most novel and load-bearing claim is the existence of a semi-congested or weakly congested regime unique to organically evolved street networks, along with the broader assertion that all three model variants exhibit a free-flow-to-congested phase transition. The reader's weakest-assumption analysis identifies the same core issue: the phase classification rests on qualitative inspection of A(t) and a single finite-time estimate of rho, without statistical or finite-size support. I agree that this is the most important threat to the central argument. The concrete test I propose directly checks whether the weakly congested states are genuine phases or long transients by extending simulation time and system size and examining whether rho(T) converges to a stable plateau. The same test framework also provides error bars for the p_c claims in the constant and degree-dependent models, which would settle whether the reported ordering of critical birth rates is meaningful. I do not see a stronger internal inconsistency: the extreme-event small-degree bias is plausible, consistent with the known random-walk result, and supported by the figures, though the missing threshold value m and absence of code/data remain reproducibility concerns. The authors themselves state that the four-network sample must be verified on larger samples, which further supports the conditional verdict. Overall, the reader's CONDITIONAL verdict is appropriate and does not need to change, but the paper should be revised to add the statistical and finite-size analysis described above before the semi-congested regime is accepted as a robust finding.","tokens_in":9740,"tokens_out":7100,"duration_ms":70326,"concrete_test":"Re-run the Mdr simulations for the points classified as weakly congested, e.g., the moderate (a0,b0) region for Delhi in Fig. 8, with simulation horizons T = 10^4, 10^5, 10^6, and 10^7 steps and with network sizes N ~ 500, 1000, and 2000 obtained by varying the city radius. Compute rho(T) = [A(T+tau) - A(T)]/(N p tau) with a fixed tau and check whether rho(T) converges to a stable positive plateau independent of T and N, decays to 0, or grows to 1. If it decays or grows, the weakly congested regime is a finite-time or finite-size artifact; if it stabilizes, the regime is real. Additionally, compute bootstrap confidence intervals for the p_c values in Figs. 4 and 5 to test whether the reported ordering pc(street) < pc(lattice) and the mild decrease of pc with eta are statistically significant.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central phase diagram, and especially the novel semi-congested regime, is established by a finite-time estimate of the order parameter rho in Eq. (2) and by visual classification of A(t) into linear, piecewise-linear, coexisting, and free-flow for selected (a0,b0) pairs (Sec. III C, Figs. 6 and 8). No statistical criterion, no tolerance for what is counted as rho = 0, no error bars, and no finite-size scaling are reported, and the simulation duration is not stated. A system that is actually in a slowly congesting transient can produce A(t) that appears piecewise-linear or shows alternating plateaus and jumps, and a finite-sample rho in (0,1), exactly the signature used to define the weakly congested regime. If those states are transient, the abstract's claim that organically evolved street networks possess a distinct semi-congested regime would not survive. The same missing statistics also weaken the p_c estimates in Figs. 4 and 5, where the statements 'pc < mu' and 'pc decreases mildly with eta' are made without confidence intervals. This concern is not that the phenomenon is impossible, but that the evidence as presented cannot distinguish a genuine phase from a long transient.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies a stochastic transport model in which particles are created at nodes with probability p, move via random hops with rejection and absorption, and are stored in FIFO queues; three parameterizations are studied: constant parameters (Model Mc), degree-dependent capacity and outflux (Model Md), and a rejection rule with scaled parameters (Model Mdr). Simulations are run on street networks of Ahmedabad, Delhi, Mumbai, and Manhattan from OpenStreetMap and on a 1000-node square lattice. The paper reports a free-flow-to-congested transition as p increases in all networks and models, a weakly congested or semi-congested regime in the organically evolved networks in Model Mdr that is absent in Manhattan and the grid, and a higher extreme-event probability at low-degree nodes than at hubs. The paper also claims that congestion cannot be fully mitigated by traffic-aware routing and that the street networks and grid behave similarly overall.","tokens_in":10001,"tokens_out":5832,"duration_ms":49602,"significance":"If the claims are correct, the paper extends previously known phase-transition and extreme-event results from synthetic scale-free networks to realistic planar street networks, and the proposed semi-congested regime in organically grown cities would be an interesting, potentially actionable phenomenon. The strengths are the use of three related transport models on real urban networks, the systematic comparison with a square lattice, and the clear demonstration that a simple random-walk-based model captures qualitatively plausible congestion behavior. The main novel claim, however, rests on a visual classification of time series without statistical error bars or finite-size scaling, and the extreme-event section omits a key threshold parameter; these need to be supplied before the results can be evaluated.","major_comments":[{"comment":"The existence and distinctness of the semi-congested/weakly congested regime is the central novel claim, but it is established by visually classifying A(t) into \"linear, piece-wise linear, coexisting and free-flow\" for selected (a0,b0) pairs and by the rho heat map in Fig. 8; no quantitative criterion, no error bars, no finite-size scaling, and no simulation length T are reported. Because a slowly congesting finite-time transient can display piecewise-linear A(t) and produce 0<rho<1, the current evidence cannot distinguish a genuine regime from a long transient. Please provide a statistical definition of the regimes, show rho as a function of observation time to demonstrate stationarity, and vary the network size (or at least state T and show error bars) for the heat-map boundaries.","section":"Sec. III C, Figs. 6-8"},{"comment":"The threshold m in qi = <ni> + m*sigma_i is never specified, and the simulation duration T in Eq. (5) is not given. Since P_EE depends on m and on the length of the stationary time series, the claim that low-degree nodes have higher extreme-event probability in Fig. 9 cannot be reproduced or quantitatively assessed. State the value of m used, report T, and show that the degree dependence is robust over a range of m.","section":"Sec. III D, Eq. (4), Fig. 9"},{"comment":"The statements \"pc < mu\" and \"pc decreases mildly with eta\" are read off from rho(p) data without defining how pc is estimated and without confidence intervals. As these statements are part of the phase-transition comparison across networks, the pc extraction rule and its uncertainty should be reported, or the claims should be formulated more cautiously.","section":"Sec. III A, Fig. 4 (and Sec. III B, Fig. 5)"}],"minor_comments":[{"comment":"The text states that a congested regime has \"rho = 0\" immediately before describing Fig. 6(a), while the same paragraph and the Fig. 6(a) caption define the congested state by rho = 1; correct this inconsistency.","section":"Sec. III C, text near Fig. 6"},{"comment":"Equations (5) and (6) are identical definitions of P_EE and should be merged into a single equation.","section":"Sec. III D, Eqs. (5) and (6)"},{"comment":"The side claim that Model Md completely eradicates congestion on scale-free networks (rho(p)=0 for any eta) is unsupported by any displayed data and is referenced to an unpublished Master's thesis [43]; this should either be shown with data or removed, since it is not needed for the paper's main argument.","section":"Sec. III B"},{"comment":"There are minor typographical errors: \"cirlce\" should be \"circle\" in the Table I caption, and \"Openstreemap\" in ref. [39] should be \"OpenStreetMap\".","section":"Table I and ref. [39]"}],"recommendation":"major_revision","confidential_remarks":"To the editor: I found the paper in scope and the problem well motivated. The main reason for major revision is not the feasibility of the claims but the evidential gap around the semi-congested regime and the missing threshold parameter m; both are addressable with additional simulation analysis rather than new theory. I would encourage the authors to add error bars and finite-size checks and to specify all simulation parameters."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read this one before you think about the abstract. The paper does something simple and useful: it runs a known traffic model (De Martino et al. with FIFO queues) on real street networks from four cities and compares to a grid. The main finding, that the free-flow to congested transition looks the same on all of these and that extreme events are more likely at small-degree nodes, is shown consistently across three model variants. That last result is a genuine extension of the scale-free network result to planar street networks, and it is not trivial.\n\nThe interesting new claim is the semi-congested regime in organically grown cities (Delhi, Mumbai, Ahmedabad) and its absence in New York and the grid. That is a plausible structural difference, and if it holds, it matters for urban traffic modeling. The problem is the evidence. The regime is defined by eyeballing A(t) shapes (linear, piecewise-linear, coexisting, free-flow) for selected (a0,b0) pairs, with no error bars, no finite-size scaling, no stated simulation time, and no threshold on what counts as rho=0. A slowly congesting transient can look exactly like piecewise-linear growth and produce rho in (0,1). The paper also never gives the value of m in the extreme event threshold, and the statements about pc<mu and the trend with eta are made without confidence intervals. None of this is fatal, but the semi-congested regime as presented is not yet a phase; it is a candidate phase.\n\nAlso minor: there is a typo in Sec. III C where the congested phase is said to have rho=0 (should be rho=1), and no code or data are provided, which makes the missing details harder to fill in. The authors do admit the four-city sample is small, which is honest.\n\nThis paper deserves a serious referee. The questions it asks are good, the simulations are cheap to reproduce in principle, and the main qualitative claims are likely to survive closer scrutiny. The revision should add error bars, a finite-size check, the actual value of m, and ideally the code. I would send it to peer review, and I would not be surprised if it comes back stronger after those fixes.","headline":"A solid simulation study with a plausible new regime that is not yet statistically established; worth sending to a referee who will demand error bars and finite-size checks.","tokens_in":10542,"tokens_out":2755,"would_cite":false,"duration_ms":22953,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["89.40.Bb","05.40.Fb"],"model":"deepseek-v4-flash","headline":"Real city street networks, like a square grid, tip from free-flow to congested traffic at a critical vehicle birth rate, and low-degree junctions host extreme events more often than hubs.","keywords":["congestion","extreme events","urban street networks","planar networks","phase transition","random walk transport model","free-flow and congested regimes"],"falsifier":"Run the same three model variants on, say, 50 real street networks of widely varying size, measuring $\\rho$ with error bars from many independent runs. If the semi-congested band ($0<\\rho<1$) disappears or shrinks as network size grows, or if any network in free flow shows hub nodes with extreme-event probability as high as low-degree nodes, the paper's universality claims fail.","tokens_in":9522,"feed_emoji":"🚦","tokens_out":11970,"duration_ms":93727,"temperature":0.7,"pith_summary":"This paper uses a random-walk model of vehicles—each node can create, serve, queue, and reject traffic—to ask whether real city street networks show the same congestion and extreme-event physics as simpler network models. Across the street networks of Ahmedabad, Delhi, Mumbai, and New York, and on a regular square grid, the authors find a free-flow-to-congested phase transition as the vehicle birth rate rises. Traffic-aware routing shifts the transition but cannot eliminate congestion. In the free-flow regime, low-degree junctions are more likely to see extreme traffic spikes than hub junctions, the same ordering previously reported for non-planar scale-free networks. The authors conclude that, for congestion and extreme events, real street networks and a square grid behave nearly alike, with a semi-congested regime appearing only in organically grown cities.","feed_headline":"City streets jam past a critical vehicle birth rate","feed_subtitle":"Same congestion transition in four cities and a grid; small junctions spike more than hubs.","key_machinery":"The load-bearing object is a transport model built from FIFO queues: at each time step a node creates a vehicle with probability $p$, serves up to $r_i$ queued vehicles into a randomly chosen neighbour, and rejects incoming traffic with probability $\\eta(n_i)$ that switches on when the queue exceeds capacity $n^*_i$. The order parameter $\\rho$ in Eq. (2) classifies the resulting state as free flow ($\\rho=0$), weakly congested ($0<\\rho<1$), or congested ($\\rho=1$). Three parameterisations of the same machinery—constant ($M_c$), degree-dependent ($M_d$), and capacity-scaled with total rejection at overload ($M_{dr}$)—probe how node degree changes the transition. Extreme events are defined within the free-flow state as crossings of the node-dependent threshold $q_i=\\langle n_i\\rangle + m\\,\\sigma_i$, and the probability $P^{\\,i}_{EE}$ is averaged over nodes of equal degree to produce the degree-ordering result.","core_discovery":"The central claim is that planar urban street networks exhibit the same congestion phase transition as a regular square grid: for all four cities and the grid, there exists a critical vehicle birth rate $p_c$ such that the total vehicle count $A(t)$ fluctuates around a mean in the free-flow phase and grows without bound in the congested phase, as measured by the order parameter $\\rho = \\lim_{t\\to\\infty}[A(t+\\tau)-A(t)]/(N p \\tau)$. This holds under three model variants—constant parameters ($M_c$), degree-dependent capacity and outflux ($M_d$), and degree-scaled outflux with total rejection at overload ($M_{dr}$). In the free-flow state, the extreme-event probability $P^{\\,i}_{EE}$, defined as the fraction of time the node's queue exceeds $q_i = \\langle n_i\\rangle + m\\,\\sigma_i$, is larger on small-degree nodes than on hubs. In the $M_{dr}$ variant, organically grown cities (Ahmedabad, Delhi, Mumbai) show a semi-congested regime with $0<\\rho<1$, where congested and free-flow behaviour coexist, while grid-like New York and the square lattice do not. The paper concludes that street networks and the square grid display similar congestion and extreme-event properties, and that the hub/low-degree extreme-event ordering does not require scale-free or non-planar topology.","pith_inferences":["If the semi-congested regime is genuine phase coexistence rather than a finite-time artifact, it should show hysteresis: ramping $p$ up and then down should trace different paths through the phase diagram. This is a testable extension the paper does not perform.","A natural next step is to ask whether the degree ordering of extreme events survives under shortest-path or congestion-aware routing, which the paper's random-walk dynamics do not include; if it fails, the ordering is tied to the dynamics rather than the planar topology.","Because low-degree junctions are the extreme-event hot spots, traffic-management schemes aimed at protecting small intersections might suppress the worst spikes more effectively than hub-focused interventions—an engineering corollary the authors do not draw."],"forward_implications":["All four city networks and the square grid show the same free-flow-to-congested transition, so a regular grid is a serviceable first model for city-wide congestion phase behaviour.","Because traffic-aware routing moves the critical birth rate but cannot remove congestion, routing alone is not a cure for network saturation.","The semi-congested regime in organically grown cities means such networks can run with some parts jammed and others free-flowing, a mixed state that grid-like New York and the square lattice do not show.","The extreme-event ordering—low-degree nodes more prone to spikes than hubs—carries over from non-planar scale-free networks to planar street networks and the square grid.","Extreme-event probabilities stay at the same order of magnitude across different $(a_0,b_0)$ pairs inside the free-flow regime, so the phenomenon does not depend on the exact parameter values."],"supporting_citations":[{"why":"Supplies the base transport model (FIFO queue, rejection probability, birth rate) that all three model variants modify.","marker":"[30, 31]"},{"why":"Establishes the small-degree-over-hubs extreme-event probability ordering on scale-free networks that this paper extends to planar street networks; also supplies the threshold definition.","marker":"[3]"},{"why":"Extends the extreme-event result to biased random walks, supporting the threshold-based event definition used here.","marker":"[4]"},{"why":"Defines the order parameter ρ used to distinguish free-flow, weakly congested, and congested regimes.","marker":"[40, 41]"},{"why":"OpenStreetMap is the public data source for the four city street networks.","marker":"[39]"},{"why":"OSMnx is the tool used to download and construct the street network graphs from OpenStreetMap.","marker":"[42]"},{"why":"Reports the Nagel-Schreckenberg result of degree-independent extreme-event probability on a small synthetic network, the counterexample the present planar-network result contrasts with.","marker":"[8]"}],"fun_headline_variants":["Critical vehicle birth rate triggers urban gridlock","Traffic jams emerge past a universal critical birth rate","Small street nodes spike more extreme traffic events","Organically grown cities show mixed congestion regime","City street congestion mirrors square grid transitions"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The claim that a distinct semi-congested regime exists rests on eyeballing the time evolution of total vehicle count for a few parameter combinations on only four cities, with no statistical test or scaling analysis to show the regime would survive with a larger network or longer simulation, and the authors themselves say more networks are needed.","fun_headline_variants_meta":{"raw":{"variants":["Critical vehicle birth rate triggers urban gridlock","Traffic jams emerge past a universal critical birth rate","Small street nodes spike more extreme traffic events","Organically grown cities show mixed congestion regime","City street congestion mirrors square grid transitions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000378,"raw_usage":{"total_tokens":2038,"prompt_tokens":999,"completion_tokens":1039,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":615,"completion_tokens_details":{"reasoning_tokens":971}},"tokens_in":615,"tokens_out":1039,"duration_ms":8673,"temperature":1.0,"reasoning_tokens":971,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T15:13:06.420349+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same three model variants on, say, 50 real street networks of widely varying size, measuring $\\rho$ with error bars from many independent runs. If the semi-congested band ($0<\\rho<1$) disappears or shrinks as network size grows, or if any network in free flow shows hub nodes with extreme-event probability as high as low-degree nodes, the paper's universality claims fail.","supporting_citations":[{"cited_title":"Kishore, M","cited_arxiv_id":null,"evidence_quote":"Establishes the small-degree-over-hubs extreme-event probability ordering on scale-free networks that this paper extends to planar street networks; also supplies the threshold definition."},{"cited_title":"Kishore, M","cited_arxiv_id":null,"evidence_quote":"Extends the extreme-event result to biased random walks, supporting the threshold-based event definition used here."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"OpenStreetMap is the public data source for the four city street networks."},{"cited_title":"Boeing, Osmnx: New methods for acquiring, con- structing, analyzing, and visualizing complex street net- works, Computers, Environment and Urban Systems65, 126 (2017)","cited_arxiv_id":null,"evidence_quote":"OSMnx is the tool used to download and construct the street network graphs from OpenStreetMap."},{"cited_title":"Gupta and M","cited_arxiv_id":null,"evidence_quote":"Reports the Nagel-Schreckenberg result of degree-independent extreme-event probability on a small synthetic network, the counterexample the present planar-network result contrasts with."}],"review_version":1}