Projected Impacts of SURTRAC-Enabled Adaptive Signal Control on Congestion and Road Safety in District 1, Ho Chi Minh City

According to data from Vietnam's Ministry of Transport and international reports as of early 2026, the country's road network exceeded 570,000 km by 2024, reflecting steady growth driven by investments under the Transport Development Strategy through 2030.

In Vietnam's logistics sector, road transport plays a pivotal role (shown in Table 1), accounting for 75–80% of freight volume (projected to reach 2.76 billion tons of goods and 9.43 billion passengers annually by 2030). This dominance arises from its flexibility compared to rail (2–3%, constrained by inadequate multimodal integration) or inland waterways (15–20%, limited by geographical factors).

Table 1. Freight transport in the ten months of 2025 by type of transport

As of September 2024, Vietnam boasted over 77 million registered motorcycles, equating to approximately 770 motorcycles/1,000 inhabitants. This number has risen from about 65 million in 2020, underscoring persistent demand for cost-effective personal mobility in both urban and rural settings. In 2025, motorcycle sales totaled 3.4 million units, reflecting a 14.9% year-on-year increase, fueled by post-pandemic economic rebound and the rapid uptake of electric two-wheelers (shown in Figure 1). This shift aligns with environmental regulations, including Hanoi's planned prohibition on internal combustion engine vehicles by mid-2026.

Figure 1. Statistics and forecasts on the number of vehicles in Vietnam [1]

Regarding four-wheeled vehicles, the registered fleet reached roughly 6.8 million units by late 2024, comprising 3.45 million vehicles with nine seats or fewer. This translates to an ownership rate of 68 vehicles per 1,000 inhabitants. Collectively, Vietnam's motorized vehicle fleet surpassed 84 million units by late 2024, with motorcycles constituting over 90%. Projections indicate sustained growth, especially in electric vehicles, potentially accounting for 30% or more of new sales by 2030, driven by urbanization and sustainability efforts.

Amid global integration and urbanization, Vietnam grapples with road crashes and congestion. In 2025, traffic incidents totaled approximately 18,500 cases, causing 10,500 fatalities and 12,200 injuries, with socioeconomic costs nearing USD 19 billion from mortality, morbidity, and healthcare. Congestion in Hanoi and Ho Chi Minh City generates USD 6–7 billion in annual losses from delays (60–90 minutes daily per commuter), fuel waste, and productivity declines, equating to 20–25% of logistics spending. Vehicle density exceeds 200% of capacity, with over 70 million motorcycles and cars growing 8–10% annually, disrupting supply chains and competitiveness amid a 2025 GDP of USD 514 billion (8.02% growth) [2].

These issues widen social inequalities, burden low-income vehicle-dependent groups, and hinder the National Transport Development Strategy to 2030 (Decision 165/2021/QĐ-TTg), targeting 5–10% annual accident reductions and congestion relief.

District 1 constitutes the administrative, commercial, financial, and tourism epicenter of Ho Chi Minh City and ranks among Vietnam’s most traffic-intensive urban districts. Although the district spans only approximately 7.7 km², it generates and attracts substantial daily traffic volumes arising from commuting patterns, commercial operations, public services, and tourism activities. Its transportation system is distinguished by exceptionally high vehicular densities, with motorcycles predominating the modal composition, while the shares of private automobiles, ride-hailing services, buses, and urban freight vehicles continue to rise.

From a traffic engineering perspective, several major arterial corridors in District 1 routinely operate at or near capacity during peak periods. Key signalized intersections, such as Ben Thanh, Phu Dong Roundabout, Me Linh Square, and the Ton Duc Thang–Nguyen Huu Canh corridor, frequently become bottlenecks. Contributing factors include elevated traffic demand, complex turning movements, and pronounced directional conflicts.

Traffic congestion in District 1 is further intensified by structural constraints inherent to its mature urban road network. Constrained roadway capacity, short inter-intersection spacing, pervasive roadside commercial activities, and frequent curbside stopping and parking collectively diminish effective roadway throughput. Under conditions approaching saturation, even minor perturbations (such as vehicle breakdowns, collisions, or unauthorized passenger pick-up and drop-off maneuvers) can precipitate network-wide congestion propagation.

The district faces elevated road safety risks resulting from frequent interactions among diverse road users, including motorcycles, private vehicles, buses, freight vehicles, and pedestrians. Conflict points are predominantly concentrated at signalized intersections, commercial precincts, and pedestrian crossing facilities. Contributing risk factors, including abrupt lane-changing behaviors, illegal stopping, improper U-turns, and pedestrian non-compliance, substantially elevate the incidence of traffic conflicts and crashes. Although severe accidents remain relatively uncommon due to prevailing low operating speeds, the prevalence of minor collisions and near-miss events is substantial. These incidents impose considerable economic costs through prolonged travel delays, elevated vehicle operating costs, increased fuel consumption, and adverse environmental externalities.

Given the severe spatial constraints precluding meaningful roadway capacity expansion in the central business district, the deployment of advanced Intelligent Transportation Systems (ITSs), AI-enabled traffic surveillance and enforcement, adaptive signal control technologies, and real-time dynamic traffic management strategies has become indispensable for mitigating congestion, bolstering road safety, and enhancing overall network operational efficiency in District 1.

Vietnam's COP26 net-zero pledge by 2050 and ASEAN sustainable transport frameworks demand advanced solutions. Thus, the study provides a scientific basis for intelligent management models, improving road logistics and sustainable development.

The distinct contribution of this study is its provision of the first projection-based, context-specific assessment of decentralized adaptive traffic control (Scalable Urban Traffic Control (SURTRAC)-type) in a Southeast Asian megacity dominated by motorcycles (85–90% of traffic volume) and heterogeneous driver behavior [3]. By developing a transparent bottom-up projection framework and a locally calibrated cost-benefit model suited to data-scarce emerging economies, the analysis bridges multi-agent traffic systems literature with practical policy evaluation in high-density urban logistics environments.

This study integrates transport-engineering mechanisms with transport-economics reasoning by treating SURTRAC’s decentralized multi-agent scheduling as a corrective instrument for classic market failures in urban road networks: namely, unpriced congestion and accident externalities that inflate logistics costs and distort resource allocation. The framework links (i) engineering-level reductions in delay variance and conflict frequency, (ii) behavioral adjustments of heterogeneous users (primarily motorcyclists), and (iii) economic outcomes measured as avoided time, accident, and emission costs within the broader logistics supply chain.

3.2 Simulated deployment scenario (2024–2025)

SURTRAC optimizes traffic signal phases by minimizing total delay, resulting in a projected 28% reduction in average vehicle delay under the projected SURTRAC scenario.

Traffic volume and delays: By 2025, volume rises to 25,000 vehicles per hour, while per-vehicle delay decreases to 54 seconds (28% reduction; daily total: 1.35 million seconds). Congestion declines from 24% to 10–15%.

Accidents and safety: Incidents are reduced by 20–25% to 90–110 annually, with damages falling by 20%.

Economic costs:

Delay cost: 28% reduction to USD 1.98 million annually:

2.75 – (2.75 × 0.28) = USD 1.98 million

Accident cost: Simulated deployment scenario, incidents are projected to decline by 20–25%, and damages by 20%, resulting in an adopted 25% cost reduction to USD 0.4125 million annually. The 20–25% reduction represents an assumed scenario derived from previous studies on adaptive traffic signal control and expert assessment and was developed through the following steps.

To derive and calibrate the projected 20–25% reduction in incidents, a surrogate safety assessment was performed using the Surrogate Safety Assessment Model (SSAM) linked to a calibrated VISSIM microscopic simulation platform. The workflow comprised the following sequential steps:

First, a base-case VISSIM model of Hang Xanh Intersection was developed and calibrated using field-measured traffic volumes (22,000 vehicles/h peak), modal composition (85% motorcycles), observed average delays (75 s), and queue length profiles. Model validation was conducted against independent delay and travel time data, achieving GEH statistics below 5.0 and speed RMSE within acceptable thresholds. Driver behaviour parameters, particularly for motorcycle following and gap acceptance, were adjusted using local trajectory data to reflect heterogeneous traffic conditions.

Second, the SURTRAC adaptive signal control logic was implemented in the validated model through external signal controller interfaces or equivalent real-time optimisation routines. Multiple simulation runs (minimum 10 random seeds) were executed for both base and SURTRAC scenarios under identical demand profiles, including the projected 2025 volume of 25,000 vehicles/h.

Third, vehicle trajectory files generated from VISSIM were processed in SSAM to identify and classify traffic conflicts. Conflict thresholds were set at Time-to-Collision (TTC) ≤ 1.5 s and Post-Encroachment Time (PET) ≤ 5.0 s, with additional filters applied to exclude non-signal-related conflicts. Conflicts were categorised by type (crossing, rear-end, lane-change) and severity to align with observed crash patterns at the intersection.

Fourth, conflict frequency and severity distributions were compared between the base and SURTRAC scenarios. Percentage reductions in total conflicts and severe conflicts (TTC < 1.0 s) were calculated. These surrogate measures were subsequently converted to expected crash reductions using locally calibrated conflict-to-crash ratios derived from the baseline video conflict study and police crash records. The resulting simulated safety improvement fell within the 18–27% range, providing quantitative support for the adopted 20–25% incident reduction interval after triangulation with literature and expert judgement.

0.55 – (0.55 × 0.25) = USD 0.4125 million

Emission cost: The projected SURTRAC scenario reduction of 25% (to approximately USD 0.4125 million) is attributed explicitly to lower idling time under the optimised signal regime. This percentage is applied uniformly to the pre-implementation emission cost base.

0.55 – (0.55 × 0.25) = USD 0.4125 million

C_post = USD 1.98 million + USD 0.4125 million + USD 0.4125 million = USD 2.805 million

Savings = C_pre – C_post = 3.85 – 2.805 = USD 1.045 million per year

Total cost under the projected SURTRAC scenario: USD 2.805 million annually (27.14% reduction; annual savings: USD 1.045 million). These quantitative estimates of costs and savings are derived from scenario assumptions and simulation-based projections. They are calibrated from U.S. SURTRAC pilot outcomes and adjusted to local traffic conditions at Hang Xanh Intersection; they do not represent measured ex-post outcomes from actual field deployment in Vietnam.

3.3 Comparative analysis and economic impacts

Delays and congestion: Pre-implementation delay averaged 75 seconds per vehicle (25% of travel time); in the projected SURTRAC scenario, it decreases to 54 seconds (28% reduction). This outcome aligns with Ministry of Transport projections for adaptive signal control, forecasting 20–30% congestion mitigation at hotspots.

Economic impacts: Annual savings of USD 1.045 million equate to 209,000 labor hours (at a USD 5/hour value of time). When scaled to a 20-intersection network in District 1 with a network-coordination multiplier of 0.15, total annual savings reach approximately USD 24.04 million.

Social and environmental impacts: Reduced accidents potentially save lives, given 380 citywide traffic fatalities in 2024, and lower emissions support Vietnam's 2030 green transport goals (Ministry of Transport).

Mechanisms of accident reduction under SURTRAC:

(i) reduced stop-go cycles and queue discharge variability, lower rear-end and sideswipe conflicts;

(ii) predictive scheduling and outflow coordination minimize dilemma-zone entries and red-light running;

(iii) smoother flow and shorter clearance intervals benefit vulnerable road users (motorcyclists and pedestrians), who constitute >85% of traffic and >70% of casualties at Ho Chi Minh City intersections.

The 20–25% reduction range is now justified by calibrating U.S. SURTRAC pilot safety outcomes and SCATS/SCOOT field studies to local conditions using a simple adjustment factor for heterogeneous traffic density. Building on these quantitative outcomes, the following section details the operating mechanism of SURTRAC.

To substantiate the economic viability of implementing the SURTRAC system in Ho Chi Minh City's District 1, this section outlines the mathematical frameworks for cost-benefit analysis. These draw on transportation economics principles, including net present value (NPV) and benefit-cost ratio (BCR), incorporating 2025 empirical data from Vietnamese urban traffic contexts.

Estimated initial capital costs: USD 500,000–1,000,000 per intersection, totaling USD 10–20 million for the district. Annual operational costs: 5–10% of capital.

5.1 Total initial capital cost calculation

The aggregate capital expenditure for SURTRAC deployment is computed as the product of the number of intersections and the per-intersection cost, incorporating hardware (e.g., sensors, radars, and processors), software integration, and installation labor. This formulation accounts for economies of scale and site-specific adaptations in Ho Chi Minh City's dense urban grid (the data calculated in this section and the calculation results are forecast data adjusted to suit local conditions).

$C_{\text{capital}}=N \times C_{\text{unit}}$               (1)

where, C_capital is the total initial capital cost (in USD), N is the number of signalized intersections (e.g., 20 in District 1, based on the simulated network topology), C_unit is the cost per intersection (ranging from 500,000 to 1,000,000 USD, calibrated from international benchmarks such as 115,810 USD per intersection for adaptive systems in United States deployments, adjusted upward by 334 to 763% for Vietnamese urban complexities including high motorcycle density and legacy infrastructure retrofitting).

5.2 Annual operational cost estimation

Ongoing expenses encompass maintenance, data processing, energy consumption, and periodic software updates, expressed as a percentage of the initial capital to reflect proportional scaling with system complexity.

$C_{\text {operational,annual }}=r \times C_{\text {capital }}$                (2)

where, C_{operational,annual} is the annual operational cost (in USD), r is the operational cost rate (0.05 to 0.10, derived from benchmarks of 10,050 USD per intersection annually in similar adaptive systems, equating to 5 to 10% of capital when aggregated across 20 intersections).

$\begin{aligned} & C_{\text {capital}}=10,000,000 \text { (USD) } C_{\text {operational,annual}}=500,000 \text { (USD) } \\ & C_{\text {capital}}=20,000,000 \text { (USD) } C_{\text {operational,annual}}=1,000,000 \text { (USD) }\end{aligned}$

5.3 Annual benefit quantification

Benefits are quantified as the sum of savings from congestion reduction and accident mitigation, prorated for District 1's contribution to citywide and nationwide impacts. Congestion benefits incorporate time savings, fuel efficiency gains, and productivity enhancements, while accident benefits include averted socio-economic losses from fatalities, injuries, and property damage.

$B_{\text {District 1,annual}}=N \times B_{\text {intersection,annual}} \times(1+\alpha)$             (3)

where, α = 0.10−0.20 is the network-coordination multiplier calibrated directly from SURTRAC U.S. pilot data.

Annual economic benefit per intersection: USD 1.045 million.

B = 20 × 1.045 × (1 + 0.15) = USD 24.035 million per year

These District 1-level annual benefits (USD 24.035 million) are obtained by scaling the single-intersection scenario projection using a network-coordination multiplier (α = 0.10–0.20). All figures in this section represent scenario assumptions and forward-looking projections calibrated to available pilot data and 2025 Vietnamese traffic parameters; they illustrate potential upper-bound outcomes and should be interpreted as indicative rather than definitive predictions of actual post-implementation performance.

Initial capital cost (20 intersections): USD 16,325,000.

Annual operating and maintenance cost: USD 500,000.

Real discount rate (r): 3%.

Analysis period (n): 10 years.

The present value (PV) of an annuity is given by:

$P V=A \times \frac{1-(1+r)^{-n}}{r}$          (4)

where, A is the annual amount. The annuity factor for r = 0.03 and n = 10 is 8.5302.

PV of benefits: PV_Benefits = 24.035 × 8.5302 = 205.02 million USD.

PV of costs: PV of operating and maintenance costs: 500,000 × 8.5302 = 4,265,100 USD

Total PV of costs: 16,325,000 + 4,265,100 = 20,590,100 USD.

5.4 Benefit-cost ratio

The BCR evaluates project efficiency by comparing discounted benefits to costs, confirming economic justification.

$B C R=\frac{P V(B)}{P V(C)}=\frac{205.02}{20.5901}=9.96: 1$               (5)

The resulting BCR of 9.96:1 is a scenario-based outcome under the optimistic calibration parameters adopted in this projection exercise. It is presented for illustrative purposes to demonstrate economic viability under stated assumptions and does not constitute an ex-ante guarantee of returns from actual deployment.

The projection-based case study at Hang Xanh Intersection, together with the illustrative cost-benefit analysis for a 20-intersection network in Ho Chi Minh City’s District 1, offers a focused and data-grounded assessment of the potential performance of SURTRAC adaptive traffic control in a high-density, motorcycle-dominated urban environment. At the intersection level, the calibrated projections indicate a 28% reduction in average per-vehicle delay (from 75 s to 54 s) and a 20–25% reduction in accident frequency, yielding annual economic savings of approximately USD 1.045 million. Scaled to the District 1 network, these benefits support a BCR of approximately 9.96:1 over 10 years.

These results illustrate the practical synergy between decentralized multi-agent scheduling and real-time traffic-flow optimization. By replacing fixed-time signals with schedule-driven, predictive control that minimizes cumulative delay while respecting safety constraints, SURTRAC can measurably reduce stop-go cycles, queue-discharge variability, and dilemma-zone conflicts at a single major intersection [19]. The District 1 cost-benefit calculations further confirm that, under the 2025 cost and benefit parameters used in the study, the intervention appears economically justified even after accounting for capital, operational, and site-specific adaptation costs typical of Vietnamese urban retrofitting.

From a policy standpoint, the findings suggest that targeted deployment of SURTRAC-type adaptive systems at high-congestion intersections in Ho Chi Minh City could contribute to measurable reductions in local delay and accident costs. Such localized improvements, if validated through field implementation, would represent one practical instrument for enhancing traffic efficiency and safety within the city’s existing road network. Policymakers may therefore consider the present analysis as supporting evidence for pilot-scale investments in intelligent traffic-signal technology, provided that subsequent empirical monitoring confirms the projected performance under real operating conditions.

Scientifically, the study advances the applied literature by developing a transparent, bottom-up projection framework calibrated to the specific challenges of motorcycle-dominated traffic and legacy infrastructure in Southeast Asian megacities. It demonstrates how decentralized multi-agent methods can be adapted to data-scarce environments without requiring full network-wide centralization. At the same time, the analysis explicitly acknowledges its limitations (reliance on forward-looking projections, secondary data for cost proration, and the exclusion of certain behavioral and institutional variables), thereby highlighting the need for controlled field trials, primary sensor-based data collection, and refined agent-based simulations to strengthen future estimates.

In summary, while the present case study does not claim nationwide or economy-wide impacts, it provides a rigorous, context-specific foundation for evaluating the technical feasibility and economic viability of SURTRAC integration at the intersection and district scales in Ho Chi Minh City. This evidence-based approach can inform more targeted, stepwise policy decisions aimed at improving urban traffic management in Vietnam’s rapidly growing megacities.