IoT-Enabled Adaptive Traffic Management: A Multiagent Framework for Urban Mobility Optimisation.

Traffic congestion and delays at signalized intersections remain a challenge in urban mobility. To minimize congestion, adaptive traffic signal control systems have been adopted across the United States. However, for these systems to function at their best, systems require an accurate estimation of traffic states, such as traffic density and queue length. This paper implements a relatively novel method, Physics-Informed Neural Networks (PINNs), for real-time traffic density estimation to enhance adaptive intersection signal control. The methodology leverages PINNs, which integrate both data-driven learning and physical constraints imposed by the Lighthill-Whitham-Richards (LWR) macroscopic traffic flow model. A simulation of a four-way intersection was used to generate realistic spatio-temporal traffic data to evaluate the performance of the model. Separate PINNs models, one trained for each incoming approach using simulation-based spatio-temporal density data, provide segment-level density predictions. These real-time density estimates are then employed within an adaptive traffic signal control algorithm designed to adjust green light phasing based on detected demand. The proposed PINNs control strategy was evaluated through simulation, comparing its performance against the Webster traffic timing method using key metrics such as average vehicle delay at the intersection and average vehicle travel time. The findings show an approximate 18% decrease in average vehicle delay and more than a 6% reduction in average travel time when utilizing adaptive traffic control based on the PINNs method for determining density values. These findings highlight the potential of leveraging PINNs for accurate, real-time traffic state estimation, which can effectively inform and improve the performance of adaptive traffic signal control systems, leading to reduced congestion and enhanced mobility at intersections.

Prelims

Start-up lost time is the time lost in the starting of the green time interval when a traffic signal phase changes from red to green and previously stopped vehicles in the curb line queue need time to accelerate to the desired speed. Actual traffic data analytics from newly installed loop coil detectors at all approaching upstream road segments of major intersections on Sathorn Road in Bangkok, Thailand, are used to confirm that the vehicle flow is obstructed considerably by the large start-up lost time. In this paper, the effect of a large start-up lost time is evaluated in terms of the travel time of passenger cars in a calibrated microscopic traffic simulation. The evaluation is based on the simulation of the urban mobility platform, while the traffic signal lights at major intersections are based on the standard Synchro optimization software. By the simulation, the average travel time per vehicle increases from 4% to 37% when the start-up lost time is varied from a baseline value of 1 s to the maximum value of 15 s, which potentially occurred in the actual traffic data collection in this paper. In addition, the optimal traffic signal green phase lengths increase from 2% to 42%, depending on the volume-to-capacity ratio. The similar increasing trend of optimal green time and average travel time per vehicle is observed using theoretical analysis based on M/M/1 and D/D/1 queues to support the results from Synchro. Findings of this paper are beneficial for understanding the impact of start-up lost time at signalized intersections.

Giving right of way green signal for the Emergency Service Vehicles (ESV) like ambulance and fire service rescue vehicle with minimum delay is a major challenge in Intelligent Transportation Systems (ITS). This challenge is currently achieved by various conventional Emergency Vehicle Pre-emption (EVP) Systems. These conventional systems fail miserably due to false triggering, line of sight issues, network delays, high cost and low reliability. In this paper, we propose a new decentralized Radio Frequency (RF) and Global Navigation Satellite System (GNSS) based Geo-fencing approach which allows smooth passage of ESVs through the junctions with minimum delays. The proposed system was field trailed in one of the junctions in Thiruvananthapuram, India. Pre and post-implementation travel time of emergency vehicles at the junction was evaluated by an external agency National Transportation Planning and Research Centre (NATPAC), Thiruvananthapuram. The evaluation result shows that there is a considerable reduction in the average travel time taken by the emergency vehicles to cross the cordon area in the post-implementation stage as compared to the pre-implementation stage. The reduction in average travel time is approximately 48%.

We propose a multi-agent simulation environment for logistics warehouses. Simulation is a crucial part of designing industrial systems, such as logistics warehouses. A warehouse is a multi-agent system (MAS) that consists of various autonomous subsystems with robots, material-handling equipment, and human workers. It is generally difficult to analyze the performance of a MAS thus, it is important to model a warehouse and conduct simulations to design and evaluate the possible system configurations. However, the cost of modeling warehouses and modifying the models is high because there are various components and interactions compared to conventional multi-agent simulations. We proposed a self-contained agent architecture and message architecture of a multi-agent simulation environment for logistics warehouses to reduce the simulation-model development and modification costs. We quantitatively evaluated our environment in terms of development costs by comparing such costs of our environment and a widely used multi-agent simulation environment.

Rapid urbanization in developing cities exacerbates transportation challenges, underscoring the critical interplay between land-use configuration and traffic flow dynamics. This study aimed at analyzing the dynamics between land-use configuration and traffic flow in Ilorin Metropolis, Nigeria, a medium-sized city experiencing acute congestion due to unplanned spatial growth, with the aim of generating evidence-based strategies for sustainable transport planning. A mixed-methods approach was utilized, combining data collection through traffic surveys at six major intersections during peak and off-peak periods, spatial analysis of satellite imagery, and 344 administered questionnaires across three Local Government Areas. Analytical techniques included descriptive statistics, Principal Component Analysis (PCA) to determine the factors shaping land-use patterns, while Multiple Linear Regression modeled the relationship between varying land-use types and traffic flow. Key findings reveal a dominant residential land-use (37.6%), with significant unplanned conversion to commercial uses along arterial corridors, yet commercial and market zones exert disproportionate impacts on traffic dynamics. PCA identified economic activities as the paramount driver of land-use patterns (81.54% variance), followed by population density (11.91%). R = 0.507, confirmed commercial areas and markets as the primary significant predictors of traffic congestion, accounting for 50.7% of traffic flow variability. Critical intersections like Muritala/Amilegbe handle over 4,000 vehicles per hour during peaks, with congestion exacerbated by inadequate infrastructure, poor signalization, and roadside trading. The study concludes that Ilorin’s traffic inefficiencies are fundamentally rooted in a misalignment between land-use concentration and transport capacity. It. recommends integrated land-use and transport policies, modernization of intersections, decentralization of commercial hubs, and investment in sustainable public transport to foster balanced urban mobility.

Despite the importance of reliability, few countries monitor it or explicitly incorporate it into transport policy making. Nevertheless, a number of policy options are available to improve this aspect of transport management. Active management of the network through ramp metering is recognized as an efficient way to control motorway traffic and improve average travel time. Far less is said about the reliability benefits of ramp metering. This paper assesses reliability benefits of ramp metering on the basis of a before-and-after study on the A6W, a French motorway near Paris. Several indicators for travel time variability are applied before and after the introduction of ramp metering. For purposes of taking reliability into account in policy impact evaluation, cost–benefit assessment provides a consistent framework within which to assess the monetized benefits. The monetary value of the reliability benefits of ramp metering is therefore calculated, and the policy implications of the results are discussed. The results suggest that in addition to providing gains in average travel time, ramp metering significantly improves reliability of travel times. It is also proposed that indices such as buffer time or planning time are useful for communicating the results, both for network operators and for users. Failing to unbundle the time-savings benefits of a project between average travel time and the variability in travel time is likely to lead to suboptimal policy solutions. Managing existing capacity better can be a cost-effective way to improve both average travel time and the variability in travel time.

Intelligent Transportation Systems (ITS) are quickly changing urban mobility in smart cities, leveraging cutting edge technologies to address critical challenges in urban transportation. This paper provides a comprehensive review of the integration of Internet of Things (IoT), Artificial Intelligence (AI), and autonomous vehicles in ITS for sustainable urban mobility. Through a systematic literature review, we analyze the latest developments in traffic flow prediction, real-time traffic control, and vehicle-to-infrastructure communication. Our findings show that IoT-enabled sensors and AI-driven models significantly improve traffic flow prediction accuracy, while autonomous and connected vehicles show promise in improving intersection management and overall system efficiency. However, critical challenges in cybersecurity, data privacy, and infrastructure readiness that must be addressed for successful implementation is identified. Analysis suggests that when properly implemented, these technologies can lead to safer, more efficient, and environmentally friendly urban transportation systems, thereby enhancing the overall sustainability of cities.

We enhance the operational performance of a textile factory by optimizing its production layout. The optimized layout improves operational time and increases overall productivity through strategic layout planning and workflow adjustments. The systematic layout planning method, combined with the analytical hierarchy process and technique for order of preference by similarity to an ideal solution from the multi-criteria decision-making process, is used to develop and select the optimal layout solution. The results show that the implementation of the improved layout design in alternative 1 led to a 45% reduction in average travel time and a 34% reduction in travel distance. This proposed solution optimizes production space while improving flexibility, safety, and competitiveness.

Provision of public transport with autonomous vehicles may enable the implementation of more flexible services, characterised by a reduction of the vehicle's capacity (and increase in the number of vehicles) — leading to fewer stops, better frequencies, and more individualised services (origins and destinations). If public transport provision is centralised, increasing flexibility will lead to reductions in average travel times (although not to the optimal capacity), as the provider internalises congestion costs; however, if public transport is provided in competitive conditions, it may lead to situations in which the entire society ends up worse off, representing a clear transportation paradox.

Background: Uncooperative driving behavior in merging areas significantly increases traffic congestion and conflicts, resulting in reduced safety and comfort of vehicle travel. With the advancement of technologies for intelligent connected environments, vehicle-to-infrastructure (V2I) technologies have been implemented to enable real-time data exchange between vehicles and infrastructure. This exchange allows vehicles to make driving strategies and motion plans based on roadside information. Therefore, it becomes particularly important to develop effective merging strategies. Various methods, both patented and non-patented, have been proposed to solve the merging challenges, aiming to optimize data exchange and enhance the formulation of driving strategies. Objective: In this paper, a strategy is proposed using roadside information to guide vehicles alternately merging, aiming at solving the safety and traffic efficiency problems in the ramp merging area, which has two main objectives: guiding vehicles to merge smoothly and safely under high traffic flow to reduce vehicle conflicts and calculating their recommended velocity through connected information to minimize congestion. Methods: A roadside system was designed to guide vehicles to merge alternately. GPS information from connected vehicles is utilized to achieve cooperative perception and localization. A time-tomerge (TTM) algorithm was developed to optimize vehicle trajectories combined with the responsibility- sensitive-safety (RSS) theory, which assigns merge sequence and recommended velocity to each vehicle based on the actual traffic conditions. Results: Simulation results indicate that this merging strategy eliminates conflicts and suppresses the formation of deceleration waves compared to the first-in-first-out (FIFO) strategy. Compared to the natural merging of independently controlled vehicles, it results in a 33.71% increase in average vehicle velocity and a 14.68% reduction in average travel time through the merge area. Conclusion: This strategy improves driving safety and efficiency by enabling vehicles to reach a common understanding of the merging sequence, and the more congested the traffic, the more significant improvement in merging efficiency.

The location of a facility can be examined in the context of several classical frameworks in economic geography and operation research where the facility is located such that some objective function is minimised (ie. Love et al., 1988; Ghosh and McLafferty, 1987; Wrigley, 1988; Fotheringham and O’Kelly, 1989). A frequently encountered objective, for example, is to find the optimal location of a new facility in terms of minimising the average time it takes individuals to travel to the nearest facility. That is, a new facility is added to an existing spatial distribution of facilities in order to achieve the maximum reduction in average travel times (the p-median problem). Other objective functions can of course be used to locate the new facility: it could for example, be located so that the maximum distance any individual has to travel is minimised (the minimax problem). Another slant on the problem is to simultaneously locate a set of facilities and to determine the allocation of demand to each of these facilities (the location-allocation problem). Still another is to model the choice of a facility by an individual as a probabilistic function of the attributes of each facility rather than as a deterministic one (a spatial interaction problem).

Leveraging large language models (LLMs) in traffic signal control (TSC) improves optimization efficiency and interpretability compared to traditional reinforcement learning (RL) methods. However, existing LLM-based approaches are limited by fixed time signal durations and are prone to hallucination errors, while RL methods lack robustness in signal timing decisions and suffer from poor generalization. To address these challenges, this paper proposes HeraldLight, a dual LLMs architecture enhanced by Herald guided prompts. The Herald Module extracts contextual information and forecasts queue lengths for each traffic phase based on real-time conditions. The first LLM, LLM-Agent, uses these forecasts to make fine grained traffic signal control, while the second LLM, LLM-Critic, refines LLM-Agent's outputs, correcting errors and hallucinations. These refined outputs are used for score-based fine-tuning to improve accuracy and robustness. Simulation experiments using CityFlow on real world datasets covering 224 intersections in Jinan (12), Hangzhou (16), and New York (196) demonstrate that HeraldLight outperforms state of the art baselines, achieving a 20.03% reduction in average travel time across all scenarios and a 10.74% reduction in average queue length on the Jinan and Hangzhou scenarios. The source code is available on GitHub: https://github.com/BUPT-ANTlab/HeraldLight.

A novel sensor network configuration is introduced for traffic surveillance in Intelligent Transportation Systems. Distributed sensor nodes based on passive acoustic transducers are non intrusively/obtrusively deployed at motorway road-side to provide real-time traffic data collection for dynamic queue/jam detection. The system is capable to provide a complete real-time picture of traffic flow at unprecedented time/space scale. The infrastructure is based on a stand-alone, install-and-forget architecture, featuring an energy efficient protocol for sensor networking and providing connectivity to back-end computing environment via TCP/IP over UMTS. Extensive experimental demonstration is given for long term operation on a motorway test site.

Modeling complex environments is a challenging problem that is compounded when there are multiple agents acting together as a team, and the team needs to maintain its own goals while allowing the agents to have some level of autonomy. We prepose a modeling framework for strategy-based multi-agent multi-team simulation environments. For these types of environments it is necessary to have a modeling infrastructure that allows for high-level, high-complexity, hierarchical interactions where team goals are prevalent but individual needs are balanced. Such modeling is proposed in this paper – modeling that will avoid large, monolithic models while maintaining complexity of expression balanced with simplicity of operation.