How do active road users act around autonomous vehicles? An inverse reinforcement learning approach

Transportation Research Part C Emerging Technologies, Journal Year: 2024, Volume and Issue: 161, P. 104572 - 104572

Published: March 16, 2024

The inevitable impact of autonomous vehicles (AV) on traffic safety is becoming a reality with the progressive deployment these in different parts world. Still, many questions linger minds road users that will share and interact AVs daily basis. To answer some questions, this study utilized recently collected real-world AV data from United States, focus mainly targeting active users. Specifically, 1,492 h recorded trips were processed to extract AV-pedestrian AV-cyclist interactions movement types. then investigated gain better understanding users' behavior, while excluding any involved intervention AVs' human test drivers. Through deep maximum entropy inverse reinforcement learning (DME-IRL), reward functions describing utility retrieved assessed for five interaction scenarios, including parallel, opposing, crossing, turning (left right) interactions. In addition, policies developed as part solution used simulate behavior validate resulting conflicts terms evasive actions. Overall, approach demonstrated high accuracy mimicking when encountering an AV, 81–84% predicting actions parallel opposing 12–17% mean absolute error indicators crossing provided reliable insight onto preferences considerations situations. cyclists tend be less cautious around compared pedestrians, slow down leave sufficient distance all most cases. robotic AVs, which can sometimes inconsistent, leads risky by users, affect other busy intersections.

Language: Английский

---

Exploring the effect of perceived safety in first/last mile mode choices

Transportation, Journal Year: 2024, Volume and Issue: unknown

Published: April 13, 2024

Abstract Micro-mobility transport modes like e-bikes and e-scooters promise higher flexibility when covering the first/last mile trip from/to public stop/station to destination point vice-versa. However, safety concerns about riding a micro vehicle in mixed traffic limit of shared mobility make conventional ones still more attractive, e.g., private car walking. This study investigates effect perceived mode choice by conducting an image-based double stated preference experiment targeted at potential micro-mobility users developing ordinal logistic regression models. The Value-of-Safety (VoS) is introduced. It refers additional distance user willing exchange avoid unsafe path. Main findings show that space can be middle-ground solution, as it reports lower heterogeneity among individuals terms perceptions. intensive use mixed-traffic decreases pedestrians, while e-bikers are threatened existence heavy motorized traffic. Low mean VoS also reported for e-scooters, demonstrating unwillingness service either detour or this vehicle. e-bike estimated almost equal car. could be, hence, concluded systematically explain unobserved disutility e-bikes.

Language: Английский

---

For whom is sharing really scaring? capturing unobserved heterogeneity in perceived comfort when cycling in shared spaces

Transportation Research Part F Traffic Psychology and Behaviour, Journal Year: 2024, Volume and Issue: 103, P. 306 - 318

Published: April 27, 2024

Language: Английский

---

Contextualizing urban road network hierarchy and its role for sustainable transport futures: A systematic literature review using bibliometric analysis and content analysis tools

Frontiers of Engineering Management, Journal Year: 2024, Volume and Issue: unknown

Published: June 18, 2024

Abstract Urban road networks play a crucial role in transport and urban planning have the potential to contribute more sustainable futures if their hierarchy is properly understood. However, concept of network hierarchy, which refers street classification prioritization, not well defined within domain engineering management, leaving many questions unanswered. Is it simply tool, or does extend defining essence cities? qualitative quantitative concept? Does emerge organically require proactive planning? Given lack comprehensive answers these questions, this research aims provide contextual understanding through lens futures. To purpose, we conducted systematic literature review, an effective method for consolidating knowledge on specific topic. A total 42 articles were analyzed using both bibliometric analysis content analysis. Our work demonstrates that consists 16 sub-concepts. Four main trends identified discussed: a) morphology structure, b) advanced algorithms classification, c) integrated planning, d) social dimension classification. Recent indicates shift toward alternative approaches prioritize mobility over car-centric models. In conclusion, our reveals multifaceted yet under researched “vehicle change,” which, utilized effectively, offers opportunities reimagine environments.

Language: Английский

---

A WebGIS Mobility Dashboard for Analyzing Spatiotemporal Traffic Flow and Public Transport Dynamics in Athens, Greece

Lecture notes in intelligent transportation and infrastructure, Journal Year: 2025, Volume and Issue: unknown, P. 886 - 899

Published: Jan. 1, 2025

Language: Английский

---

A framework for analyzing driver safety-efficiency trade-offs at uncontrolled crosswalks: Towards social vehicle automation

Safety Science, Journal Year: 2025, Volume and Issue: 187, P. 106860 - 106860

Published: March 26, 2025

Language: Английский

---

Between Looms and Beds: Unveiling Transportation Challenges with a Radius-Led Mixed-Methods Approach in Informal Space Based on a Study Conducted in Guangzhou, China

Buildings, Journal Year: 2025, Volume and Issue: 15(7), P. 1185 - 1185

Published: April 4, 2025

We investigate the conflicts between formal and informal urban spaces how policies’ neglect of needs exacerbates traffic chaos segregation in East Asia, aiming to decipher operational logic transportation systems their dynamic interactions with urbanization processes. Focusing on Zhongda Textile City, we delve into specific manifestation these conflicts, which appear four key aspects: (1) mismatch planning needs, (2) physical disconnection areas, (3) infrastructure projects occupying spaces, (4) policy-making neglecting existing experiences. Using a mixed-method framework, highlight marginalization through evolving relationship provide insights strategies that account for symbiotic yet contested dynamics.

Language: Английский

---

The Unsustainable Proximity Paradox in Medium-Sized Cities: A Qualitative Study on User Perceptions of Mobility Policies

Land, Journal Year: 2025, Volume and Issue: 14(5), P. 944 - 944

Published: April 27, 2025

Medium-sized cities face unique challenges in fostering sustainable mobility due to their socio-spatial characteristics, including recent decentralized services and urban sprawl. This study examines user-centric factors influencing behaviors Caceres, Spain, through qualitative focus group analysis with 18 participants across two age groups. By employing a co-occurrence methodology, this research identifies key relationships within four thematic areas: public transport, active mobility, innovation, planning. The findings reveal persistent car dependency despite policies, driven by the following: (1) inadequate transport coordination between regional areas, poor information availability, lack of service synchronization; (2) perceived safety concerns, insufficient infrastructure for cycling, ineffective pedestrianization strategies; (3) limited adoption technological solutions cultural barriers, preference informal arrangements, usability issues apps; (4) mismatches form distribution, proximity perception, consumer preferences reinforcing dependency. underscores need integrated systems, mixed land-use planning, improved accessibility measures achieve equitable transitions. conclusion includes series policy recommendations.

Language: Английский

---

Video-Based Safety Evaluation of Same-Direction Interactions Between Cyclists and Motorized Vehicles

Transportation research procedia, Journal Year: 2025, Volume and Issue: 86, P. 435 - 442

Published: Jan. 1, 2025

Language: Английский

---

Cooperative adaptable lanes for safer shared space and improved mixed-traffic flow

Transportation Research Part C Emerging Technologies, Journal Year: 2024, Volume and Issue: 166, P. 104748 - 104748

Published: July 26, 2024

With the rapid increase in percentage of world's population living cities, design existing transportation infrastructure requires serious consideration. Current road networks, especially large face acute pressures due to increased demand for vehicles, cyclists, and pedestrians. Although much attention has been given improve traffic management accommodate via coordinating optimizing signals, research focused on adapting static allocation street spaces right-of-way dynamically based mixed flow is still scarce. This paper proposes a multi-agent reinforcement learning (RL) agent approach that cooperatively adapts individual lane widths access permissions real-world flow. In particular, multiple cooperative agents are trained with temporal data learn decide suitable motorized bicycles, pedestrians, along whether co-sharing space between pedestrians cyclists safe. Using microscopic simulator model four-legged intersection, we our RL synthetic data, tested it realistic multi-modal data. The proposed reduces overall average waiting time queue length by 48.9% 37.7%, respectively, compared Static (baseline) design. Additionally, observe CALM's ability gradually adjust widths, contrasting Heuristic implementation's erratic adjustments, which pose potential safety concerns. Notably, learns adaptively toggle as one co-shared lane, ensuring comfort maintaining level service according designer's policy. Finally, demonstrate scalability simulated large-scale network.

Language: Английский

---