arXiv subjects

Rafał Kucharski

Publications and source records attributed to Rafał Kucharski.

At least 19 recordsLinked to original sources

arXiv2026

Randomized routing strategies of fleets of CAVs may prove market efficient

In future cities every driver may own a vehicle which could be either independently driven (HDV), or autonomously routed and piloted (CAV). The autonomous operations could be handled by a few competing companies. What is the market structure which would make this market aligned with city goals? In this paper we discuss a variant of the emerging market of collectively routed fleets of CAVs, where revenue for fleet operators is proportional to market share. We provide benchmark scenarios to compare the routing algorithms. We present several routing algorithms and demonstrate that, when the attitudes of human drivers towards CAVs exhibit significant diversity, randomised CAV routing, resulting in unpredictable travel times for HDVs, is more efficient than routing proportional to system optimum/user equilibrium. Based on this, we propose to improve the design of the market by augmenting the market-share objective with mean systemwide travel time in order to limit antisocial randomised strategies of fleet operators and drive the competition towards social welfare oriented cooperation.

cs.MA↗

arXiv2026

Demand-agnostic assessment of on-demand pooled transit services

This study proposes a method to assess the potential of pooled on-demand transit feeder services in urban areas where demand is not yet known. We introduce the fraction of demand, reflecting the probability that a resident will use the service. Demand is generated on the distribution of residents address points at varying demand fraction levels. Through simulations, we match travellers into pooled rides and evaluate the service potential using three performance indicators (KPIs). We observe how these KPIs change with varying demand fractions and identify the most promising hub for each area. By setting KPI thresholds, we select the optimal combination of area and hub that meets these thresholds at the lowest demand fraction. This approach provides municipalities with a structured tool for pre-deployment evaluation, helping them choose the most suitable areas for launching new services despite the absence of exact demand data. We illustrate its application through a case study in Krakow, ranking 12 pre-selected areas for feeder service deployment.

physics.soc-ph↗

arXiv2026

PC3D: Zero-Shot Cooperation Across Variable Rosters via Personalized Context Distillation

Cooperative multi-agent reinforcement learning often assumes a fixed execution team, yet many decentralized systems must operate with varying numbers of active agents during deployment. We study this setting under episodic roster variation: each episode is executed by a set of homogeneous agents, with the team size varying across episodes. Agents act only from local histories, without execution-time communication, privileged coordinators, or online retraining. Therefore, effective cooperation requires each agent to recover relevant context about the active team and adapt its behavior accordingly. To this end, we propose PC3D (Personalized Central Coordination Context Distillation), a method for training decentralized policies to recover and use personalized coordination context from local interaction histories. During training, a set-structured centralized teacher compresses the active team into coordination tokens and personalizes them into agent-specific contexts, which are distilled into decentralized policies. At execution, each agent predicts its own context from local history and adaptively uses it to condition decision-making. Across three cooperative MARL benchmarks, PC3D achieves higher returns than the evaluated baselines with both seen and unseen roster sizes, and ablations attribute these gains to both context distillation and adaptive context use.

cs.LG↗

arXiv2026

Generalising Travel Time Prediction To Varying Route Choices In Urban Networks

Previous methods that predict system-wide travel time, predominantly grounded in graph neural networks, remain limited to typical and recurring demand patterns. While they successfully predict future congestion following daily commute, they inherently approximate a single demand realisation and fail to capture varying route choices. In this work, we propose a Generalised Travel Time Predictor (GenTTP) that successfully differentiates route choices and offers accurate flow and travel time predictions. Our framework learns to uncover complex spatiotemporal traffic patterns and microscopic relationships between route choices and the resulting travel times. This addresses a critical gap: the lack of travel time prediction models that generalise across varying route assignments, where the same demand can produce substantially different network-wide outcomes depending on how travellers are distributed over available paths.

cs.MA↗

arXiv2025

Market share maximizing strategies of CAV fleet operators may cause chaos in our cities

We study the dynamics and equilibria of a new kind of routing games, where players - drivers of future autonomous vehicles - may switch between individual (HDV) and collective (CAV) routing. In individual routing, just like today, drivers select routes minimizing expected travel costs, whereas in collective routing an operator centrally assigns vehicles to routes. The utility is then the average experienced travel time discounted with individually perceived attractiveness of automated driving. The market share maximising strategy amounts to offering utility greater than for individual routing to as many drivers as possible. Our theoretical contribution consists in developing a rigorous mathematical framework of individualized collective routing and studying algorithms which fleets of CAVs may use for their market-share optimization. We also define bi-level CAV - HDV equilibria and derive conditions which link the potential marketing behaviour of CAVs to the behavioural profile of the human population. Practically, we find that the fleet operator may often be able to equilibrate at full market share by simply mimicking the choices HDVs would make. In more realistic heterogenous human population settings, however, we discover that the market-share maximizing fleet controller should use highly variable mixed strategies as a means to attract or retain customers. The reason is that in mixed routing the powerful group player can control which vehicles are routed via congested and uncongested alternatives. The congestion pattern generated by CAVs is, however, not known to HDVs before departure and so HDVs cannot select faster routes and face huge uncertainty whichever alternative they choose. Consequently, mixed market-share maximising fleet strategies resulting in unpredictable day-to-day driving conditions may, alarmingly, become pervasive in our future cities.

math.OC↗

arXiv2025

URB -- Urban Routing Benchmark for RL-equipped Connected Autonomous Vehicles

Connected Autonomous Vehicles (CAVs) promise to reduce congestion in future urban networks, potentially by optimizing their routing decisions. Unlike for human drivers, these decisions can be made with collective, data-driven policies, developed using machine learning algorithms. Reinforcement learning (RL) can facilitate the development of such collective routing strategies, yet standardized and realistic benchmarks are missing. To that end, we present URB: Urban Routing Benchmark for RL-equipped Connected Autonomous Vehicles. URB is a comprehensive benchmarking environment that unifies evaluation across 29 real-world traffic networks paired with realistic demand patterns. URB comes with a catalog of predefined tasks, multi-agent RL (MARL) algorithm implementations, three baseline methods, domain-specific performance metrics, and a modular configuration scheme. Our results show that, despite the lengthy and costly training, state-of-the-art MARL algorithms rarely outperformed humans. The experimental results reported in this paper initiate the first leaderboard for MARL in large-scale urban routing optimization. They reveal that current approaches struggle to scale, emphasizing the urgent need for advancements in this domain.

cs.LG↗

arXiv2025

Autonomous vehicles need social awareness to find optima in multi-agent reinforcement learning routing games

Previous work has shown that when multiple selfish Autonomous Vehicles (AVs) are introduced to future cities and start learning optimal routing strategies using Multi-Agent Reinforcement Learning (MARL), they may destabilize traffic systems, as they would require a significant amount of time to converge to the optimal solution, equivalent to years of real-world commuting. We demonstrate that moving beyond the selfish component in the reward significantly relieves this issue. If each AV, apart from minimizing its own travel time, aims to reduce its impact on the system, this will be beneficial not only for the system-wide performance but also for each individual player in this routing game. By introducing an intrinsic reward signal based on the marginal cost matrix, we significantly reduce training time and achieve convergence more reliably. Marginal cost quantifies the impact of each individual action (route-choice) on the system (total travel time). Including it as one of the components of the reward can reduce the degree of non-stationarity by aligning agents' objectives. Notably, the proposed counterfactual formulation preserves the system's equilibria and avoids oscillations. Our experiments show that training MARL algorithms with our novel reward formulation enables the agents to converge to the optimal solution, whereas the baseline algorithms fail to do so. We show these effects in both a toy network and the real-world network of Saint-Arnoult. Our results optimistically indicate that social awareness (i.e., including marginal costs in routing decisions) improves both the system-wide and individual performance of future urban systems with AVs.

cs.MA↗

arXiv2025

Equilibria in routing games with connected autonomous vehicles will not be strong, as exclusive clubs may form

User Equilibrium is the standard representation of the so-called routing game in which drivers adjust their route choices to arrive at their destinations as fast as possible. Asking whether this Equilibrium is strong or not was meaningless for human drivers who did not form coalitions due to technical and behavioral constraints. This is no longer the case for connected autonomous vehicles (CAVs), which will be able to communicate and collaborate to jointly form routing coalitions. We demonstrate this for the first time on a carefully designed toy-network example, where a `club` of three autonomous vehicles jointly decides to deviate from the user equilibrium and benefit (arrive faster). The formation of such a club has negative consequences for other users, who are not invited to join it and now travel longer, and for the system, making it suboptimal and disequilibrated, which triggers adaptation dynamics. This discovery has profound implications for the future of our cities. We demonstrate that, if not prevented, CAV operators may intentionally disequilibrate traffic systems from their classic Nash equilibria, benefiting their own users and imposing costs on others. These findings suggest the possible emergence of an exclusive CAV elite, from which human-driven vehicles and non-coalition members may be excluded, potentially leading to systematically longer travel times for those outside the coalition, which would be harmful for the equity of public road networks.

cs.GT↗

arXiv2025

Impact of Collective Behaviors of Autonomous Vehicles on Urban Traffic Dynamics: A Multi-Agent Reinforcement Learning Approach

This study examines the potential impact of reinforcement learning (RL)-enabled autonomous vehicles (AV) on urban traffic flow in a mixed traffic environment. We focus on a simplified day-to-day route choice problem in a multi-agent setting. We consider a city network where human drivers travel through their chosen routes to reach their destinations in minimum travel time. Then, we convert one-third of the population into AVs, which are RL agents employing Deep Q-learning algorithm. We define a set of optimization targets, or as we call them behaviors, namely selfish, collaborative, competitive, social, altruistic, and malicious. We impose a selected behavior on AVs through their rewards. We run our simulations using our in-house developed RL framework PARCOUR. Our simulations reveal that AVs optimize their travel times by up to 5\%, with varying impacts on human drivers' travel times depending on the AV behavior. In all cases where AVs adopt a self-serving behavior, they achieve shorter travel times than human drivers. Our findings highlight the complexity differences in learning tasks of each target behavior. We demonstrate that the multi-agent RL setting is applicable for collective routing on traffic networks, though their impact on coexisting parties greatly varies with the behaviors adopted.

cs.MA↗

arXiv2025

Wardropian Cycles make traffic assignment both optimal and fair by eliminating price-of-anarchy with Cyclical User Equilibrium for compliant connected autonomous vehicles

Connected and Autonomous Vehicles (CAVs) open the possibility for centralised routing with full compliance, making System Optimal traffic assignment attainable. However, as System Optimum makes some drivers better off than others, voluntary acceptance seems dubious. To overcome this issue, we propose a new concept of Wardropian cycles, which, in contrast to previous utopian visions, makes the assignment fair on top of being optimal, which amounts to satisfaction of both Wardrop's principles. Such cycles, represented as sequences of permutations to the daily assignment matrices, always exist and equalise, after a limited number of days, average travel times among travellers (like in User Equilibrium) while preserving everyday optimality of path flows (like in System Optimum). We propose exact methods to compute such cycles and reduce their length and within-cycle inconvenience to the users. As identification of optimal cycles turns out to be NP-hard in many aspects, we introduce a greedy heuristic efficiently approximating the optimal solution. Finally, we introduce and discuss a new paradigm of Cyclical User Equilibrium, which ensures stability of optimal Wardropian Cycles under unilateral deviations. We complement our theoretical study with large-scale simulations. In Barcelona, 670 vehicle-hours of Price-of-Anarchy are eliminated using cycles with a median length of 11 days-though 5% of cycles exceed 90 days. However, in Berlin, just five days of applying the greedy assignment rule significantly reduces initial inequity. In Barcelona, Anaheim, and Sioux Falls, less than 7% of the initial inequity remains after 10 days, demonstrating the effectiveness of this approach in improving traffic performance with more ubiquitous social acceptability.

eess.SY↗

arXiv2025

Detection of coordinated fleet vehicles in route choice urban games. Part I. Inverse fleet assignment theory

Detection of collectively routing fleets of vehicles in future urban systems may become important for the management of traffic, as such routing may destabilize urban networks leading to deterioration of driving conditions. Accordingly, in this paper we discuss the question whether it is possible to determine the flow of fleet vehicles on all routes given the fleet size and behaviour as well as the combined total flow of fleet and non-fleet vehicles on every route. We prove that the answer to this Inverse Fleet Assignment Problem is 'yes' for myopic fleet strategies which are more 'selfish' than 'altruistic', and 'no' otherwise, under mild assumptions on route/link performance functions. To reach these conclusions we introduce the forward fleet assignment operator and study its properties, proving that it is invertible for 'bad' objectives of fleet controllers. We also discuss the challenges of implementing myopic fleet routing in the real world and compare it to Stackelberg and Nash routing. Finally, we show that optimal Stackelberg fleet routing could involve highly variable mixed strategies in some scenarios, which would likely cause chaos in the traffic network.

math.OC↗

arXiv2025

Collaboration Between the City and Machine Learning Community is Crucial to Efficient Autonomous Vehicles Routing

Autonomous vehicles (AVs), possibly using Multi-Agent Reinforcement Learning (MARL) for simultaneous route optimization, may destabilize traffic networks, with human drivers potentially experiencing longer travel times. We study this interaction by simulating human drivers and AVs. Our experiments with standard MARL algorithms reveal that, both in simplified and complex networks, policies often fail to converge to an optimal solution or require long training periods. This problem is amplified by the fact that we cannot rely entirely on simulated training, as there are no accurate models of human routing behavior. At the same time, real-world training in cities risks destabilizing urban traffic systems, increasing externalities, such as $CO_2$ emissions, and introducing non-stationarity as human drivers will adapt unpredictably to AV behaviors. In this position paper, we argue that city authorities must collaborate with the ML community to monitor and critically evaluate the routing algorithms proposed by car companies toward fair and system-efficient routing algorithms and regulatory standards.

cs.MA↗

arXiv2025

SimFLEX: a methodology for comparative analysis of urban areas for implementing new on-demand feeder bus services

On-demand feeder bus services present an innovative solution to urban mobility challenges, yet their success depends on thorough assessment and strategic planning. Despite their potential, a comprehensive framework for evaluating feasibility and identifying suitable service areas remains underdeveloped. Simulation Framework for Feeder Location Evaluation (SimFLEX) uses spatial, demographic, and transport-specific data to run microsimulations and compute key performance indicators (KPIs), including service attractiveness, waiting time reduction, and added value. SimFLEX employs multiple replications to estimate demand and mode choices and integrates OpenTripPlanner (OTP) for public transport routing and ExMAS for calculating shared trip attributes and KPIs. For each demand scenario, we model the traveler learning process using the method of successive averages (MSA), stabilizing the system. After stabilization, we calculate KPIs for comparative and sensitivity analyzes. We applied SimFLEX to compare two remote urban areas in Krakow, Poland - Bronowice and Skotniki - the candidates for service launch. Our analysis revealed notable differences between analyzed areas: Skotniki exhibited higher service attractiveness (up to 30%) and added value (up to 7%), while Bronowice showed greater potential for reducing waiting times (by nearly 77%). To assess the reliability of our model output, we conducted a sensitivity analysis across a range of alternative-specific constants (ASC). The results consistently confirmed Skotniki as the superior candidate for service implementation. SimFLEX can be instrumental for policymakers to estimate new service performance in the considered area, publicly available and applicable to various use cases. It can integrate alternative models and approaches, making it a versatile tool for policymakers and urban planners to enhance urban mobility.

physics.soc-ph↗

arXiv2025

RouteRL: Multi-agent reinforcement learning framework for urban route choice with autonomous vehicles

RouteRL is a novel framework that integrates multi-agent reinforcement learning (MARL) with a microscopic traffic simulation, facilitating the testing and development of efficient route choice strategies for autonomous vehicles (AVs). The proposed framework simulates the daily route choices of driver agents in a city, including two types: human drivers, emulated using behavioral route choice models, and AVs, modeled as MARL agents optimizing their policies for a predefined objective. RouteRL aims to advance research in MARL, transport modeling, and human-AI interaction for transportation applications. This study presents a technical report on RouteRL, outlines its potential research contributions, and showcases its impact via illustrative examples.

cs.MA↗

arXiv2024

Social impact of CAVs -- coexistence of machines and humans in the context of route choice

Suppose in a stable urban traffic system populated only by human driven vehicles (HDVs), a given proportion (e.g. 10%) is replaced by a fleet of Connected and Autonomous Vehicles (CAVs), which share information and pursue a collective goal. Suppose these vehicles are centrally coordinated and differ from HDVs only by their collective capacities allowing them to make more efficient routing decisions before the travel on a given day begins. Suppose there is a choice between two routes and every day each driver makes a decision which route to take. Human drivers maximize their utility. CAVs might optimize different goals, such as the total travel time of the fleet. We show that in this plausible futuristic setting, the strategy CAVs are allowed to adopt may result in human drivers either benefitting or being systematically disadvantaged and urban networks becoming more or less optimal. Consequently, some regulatory measures might become indispensable.

cs.MA↗

arXiv2023

Dynamics of the Ride-Sourcing Market: A Coevolutionary Model of Competition between Two-Sided Mobility Platforms

There is a fierce competition between two-sided mobility platforms (e.g., Uber and Lyft) fueled by massive subsidies, yet the underlying dynamics and interactions between the competing plat-forms are largely unknown. These platforms rely on the cross-side network effects to grow, they need to attract agents from both sides to kick-off: travellers are needed for drivers and drivers are needed for travellers. We use our coevolutionary model featured by the S-shaped learning curves to simulate the day-to-day dynamics of the ride-sourcing market at the microscopic level. We run three scenarios to illustrate the possible equilibria in the market. Our results underline how the correlation inside the ride-sourcing nest of the agents choice set significantly affects the plat-forms' market shares. While late entry to the market decreases the chance of platform success and possibly results in "winner-takes-all", heavy subsidies can keep the new platform in competition giving rise to "market sharing" regime.

cs.MA↗

arXiv2022

Modelling the Rise and Fall of Two-Sided Mobility Markets with Microsimulation

In this paper, we propose a novel modelling framework to reproduce the market entry strategies for two-sided mobility platforms. In the MaaSSim agent-based simulator, we develop a co-evolutionary model to represent day-to-day dynamics of the two-sided mobility market with agents making rational decisions to maximize their perceived utility. Participation probability of agents depends on utility, composed of: experience, word of mouth and marketing components adjusted by agents every day with the novel S-shaped formulas - better suited (in our opinion) to reproduce market entry dynamics than previous approaches. With such a rich representation, we can realistically model a variety of market entry strategies and create significant network effects to reproduce the rise and fall of two-side mobility platforms. To illustrate model capabilities, we simulate a 400-day evolution of 200 drivers and 2000 travelers on a road-network of Amsterdam. We design a six-stage market entry strategy with consecutive: kick-off, discount, launch, growth, maturity and greed stages. After 25 days the platform offers discounts, yet it starts gaining market share only when the marketing campaign launches at day 50. Campaign finishes after 50 days, which does not stop the growth, now fueled mainly with a positive word of mouth effect and experiences. The platform ends discounts after 200 days and reaches the steady maturity period, after which its greedy strategy leads to collapse of its market share and profit. All above simulated with a single behavioral model, which well reproduces how agents of both sides adapts to platform actions.

cs.MA↗

arXiv2022

Can we start sharing our rides again? The postpandemic ride-pooling market

Before the pandemic ride-pooling was a promising emerging mode in urban mobility. It started reaching the critical mass with a growing number of service providers and the increasing number of travellers (needed to ensure ride-pooling efficiency and sustainability). However, the COVID pandemic was disruptive for ride-pooling. Many services were cancelled, several operators needed to change their business models and travellers started avoiding those services. In the postpandemic period, we need to understand what is the future of ride-pooling: whether the ride-pooling system can recover and remain a relevant part of future mobility. Here we provide an overview of the postpandemic ride-pooling market based on the analysis of three components: a) literature review, b) empirical pooling availability survey and c) travellers' behaviour studies. We conclude that the core elements of the ride-pooling business model were not affected by the pandemic. It remains a promising option for all the parties involved, with a great potential to become attractive for travellers, drivers, TNC platforms and policymakers. The travel behaviour changes due to the pandemic seem not to be long-lasting, our virus awareness is no anymore the key concern and our willingness to share and reduce fares seem to be high again. Yet, whether ride-pooling will get another chance to grow remains open. The number of launches of ride-pooling start-ups is unprecedented, yet the financial perspectives are unclear.

physics.soc-ph↗