Transportation Research Part F-Traffic Psychology and Behaviour最新文献_第3页

Predicting driver retention in ride-hailing platforms: an empirical study of socio-demographic and behavioral influences 预测网约车平台的司机留存：社会人口统计学和行为影响的实证研究

IF 4.4 2区工程技术 Q1 PSYCHOLOGY, APPLIED

Transportation Research Part F-Traffic Psychology and Behaviour

Pub Date : 2026-01-13 DOI: 10.1016/j.trf.2026.103513

Sourav Shome, Md Tawfiq Sarwar, Sourav Sarker, Chandra Shekhor Mondal

The rapid growth of ride-hailing services has transformed urban mobility in Dhaka, offering flexible transportation alternatives while creating new employment opportunities for drivers. However, the sustainability of this employment model remains uncertain, as evidenced by the recent closures of popular ride-hailing platforms like Shohoz Ride, which struggled to maintain operations despite early success. This study aims to investigate the factors that influence drivers' decisions to remain in the ride-hailing sector in Dhaka. Using ordered logistic regression analysis, this research examines the role of socio-economic, demographic, and built environmental variables—including education level, marital status, years of experience, vehicle ownership, comfort with road conditions, the impact of traffic congestion and waiting time – on driver retention. The findings reveal that drivers with higher education, urban backgrounds, and personal vehicles are less likely to stay in the industry, as they often have access to more stable and lucrative job opportunities. In contrast, drivers with more experience and rental vehicles, as well as who are comfortable navigating Dhaka's traffic conditions, exhibit a higher likelihood of continuing in the sector. Furthermore, the study highlights the significant impact of Dhaka's severe traffic congestion, which reduces the number of trips drivers can complete and adversely affects their earnings and job satisfaction. Based on these findings, the study recommends policy interventions such as improving road infrastructure, developing waiting zones, providing incentives for experienced and rental vehicle drivers, and offering greater flexibility in ride-hailing models. By addressing these challenges, policymakers and ride-hailing companies can improve driver retention, promote industry sustainability, and contribute to the long-term growth of Dhaka's ride-hailing sector.

网约车服务的快速增长改变了达卡的城市交通，提供了灵活的交通选择，同时为司机创造了新的就业机会。然而，这种就业模式的可持续性仍然不确定，最近一些受欢迎的叫车平台（如Shohoz Ride）的关闭证明了这一点，尽管早期取得了成功，但这些平台仍在努力维持运营。本研究旨在调查影响达卡司机决定留在网约车行业的因素。运用有序逻辑回归分析，本研究考察了社会经济、人口统计和构建的环境变量（包括教育水平、婚姻状况、经验年限、车辆拥有量、对路况的舒适度、交通拥堵和等待时间的影响）对驾驶员留存率的作用。调查结果显示，受过高等教育、城市背景和拥有私家车的司机不太可能留在这个行业，因为他们通常可以获得更稳定、更赚钱的工作机会。相比之下，拥有更多经验和租赁车辆的司机，以及能够自如驾驭达卡的交通状况的司机，更有可能继续从事该行业。此外，该研究强调了达卡严重的交通拥堵的重大影响，这减少了司机可以完成的行程数量，并对他们的收入和工作满意度产生了不利影响。基于这些发现，该研究建议采取政策干预措施，如改善道路基础设施、开发等候区、为有经验的司机和租赁车辆司机提供激励，以及在叫车模式中提供更大的灵活性。通过应对这些挑战，政策制定者和网约车公司可以提高司机的保留率，促进行业的可持续性，并为达卡网约车行业的长期增长做出贡献。

{"title":"Predicting driver retention in ride-hailing platforms: an empirical study of socio-demographic and behavioral influences","authors":"Sourav Shome, Md Tawfiq Sarwar, Sourav Sarker, Chandra Shekhor Mondal","doi":"10.1016/j.trf.2026.103513","DOIUrl":"10.1016/j.trf.2026.103513","url":null,"abstract":"<div><div>The rapid growth of ride-hailing services has transformed urban mobility in Dhaka, offering flexible transportation alternatives while creating new employment opportunities for drivers. However, the sustainability of this employment model remains uncertain, as evidenced by the recent closures of popular ride-hailing platforms like Shohoz Ride, which struggled to maintain operations despite early success. This study aims to investigate the factors that influence drivers' decisions to remain in the ride-hailing sector in Dhaka. Using ordered logistic regression analysis, this research examines the role of socio-economic, demographic, and built environmental variables—including education level, marital status, years of experience, vehicle ownership, comfort with road conditions, the impact of traffic congestion and waiting time – on driver retention. The findings reveal that drivers with higher education, urban backgrounds, and personal vehicles are less likely to stay in the industry, as they often have access to more stable and lucrative job opportunities. In contrast, drivers with more experience and rental vehicles, as well as who are comfortable navigating Dhaka's traffic conditions, exhibit a higher likelihood of continuing in the sector. Furthermore, the study highlights the significant impact of Dhaka's severe traffic congestion, which reduces the number of trips drivers can complete and adversely affects their earnings and job satisfaction. Based on these findings, the study recommends policy interventions such as improving road infrastructure, developing waiting zones, providing incentives for experienced and rental vehicle drivers, and offering greater flexibility in ride-hailing models. By addressing these challenges, policymakers and ride-hailing companies can improve driver retention, promote industry sustainability, and contribute to the long-term growth of Dhaka's ride-hailing sector.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"118 ","pages":"Article 103513"},"PeriodicalIF":4.4,"publicationDate":"2026-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145978289","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

The central role of threatened experiences in determining pedestrians' perceived safety 威胁体验在决定行人感知安全中的核心作用

IF 4.4 2区工程技术 Q1 PSYCHOLOGY, APPLIED

Transportation Research Part F-Traffic Psychology and Behaviour

Pub Date : 2026-01-12 DOI: 10.1016/j.trf.2026.103511

Kyu Ri Kim, Jennifer Dill

Understanding pedestrians' perceived safety and attitudes toward safety is essential for developing proactive transportation safety plans, as perceptions and attitudes can influence subsequent behaviors. Since little research has thoroughly examined threatened experiences that may affect perceived safety, this study explores the relationship between pedestrians' threatened experiences—shaped by external factors and individual characteristics—and their perceived safety. Data are from ten study areas in Oregon, USA, including surveys of individuals (n = 551) about walking behaviors, experiences of feeling threatened, and individual characteristics, as well as data on factors related to pedestrian risk, particularly crashes, within a half-mile buffer around the respondents' addresses. We examined the relationships among these factors by estimating structural equation models. The most significant factor affecting perceived safety is pedestrians' threatened experiences due to motorists' behaviors, followed by threatened experiences due to facilities. Vehicle volume (proxy), along with the number of all types of crashes, age, gender, and walking frequency, influences perceived safety, fully mediated by threatened experiences. On the other hand, the size of parks and the presence of disabilities directly affect perceived safety, independent of prior experiences. These findings suggest that changing pedestrians' perceived safety may be challenging without addressing their experiences of being threatened by drivers' inappropriate or aggressive behaviors. Therefore, transportation planning and facility design that aim to reduce driving errors or encourage safer driving are essential for improving pedestrians' experiences, which in turn positively influence their perceived safety.

了解行人对安全的感知和对安全的态度对于制定主动交通安全计划至关重要，因为感知和态度会影响随后的行为。由于很少有研究深入研究可能影响感知安全的威胁体验，因此本研究探讨了行人受到外部因素和个人特征影响的威胁体验与他们感知安全之间的关系。数据来自美国俄勒冈州的十个研究区域，包括对个人（n = 551）的调查，包括步行行为、感觉受到威胁的经历和个人特征，以及与行人风险相关的因素的数据，特别是在受访者地址周围半英里缓冲区内的撞车事故。我们通过估计结构方程模型来检验这些因素之间的关系。影响感知安全最显著的因素是行人因驾驶人的行为而产生的威胁体验，其次是设施造成的威胁体验。车辆数量（代理）与所有类型碰撞的数量、年龄、性别和步行频率一起影响感知安全，完全由威胁经历介导。另一方面，公园的大小和残疾的存在直接影响感知安全，独立于先前的经验。这些发现表明，如果不解决行人受到司机不当或攻击性行为威胁的经历，改变行人对安全的感知可能是具有挑战性的。因此，旨在减少驾驶失误或鼓励更安全驾驶的交通规划和设施设计对于改善行人的体验至关重要，这反过来又对他们的感知安全产生积极影响。

{"title":"The central role of threatened experiences in determining pedestrians' perceived safety","authors":"Kyu Ri Kim, Jennifer Dill","doi":"10.1016/j.trf.2026.103511","DOIUrl":"10.1016/j.trf.2026.103511","url":null,"abstract":"<div><div>Understanding pedestrians' perceived safety and attitudes toward safety is essential for developing proactive transportation safety plans, as perceptions and attitudes can influence subsequent behaviors. Since little research has thoroughly examined threatened experiences that may affect perceived safety, this study explores the relationship between pedestrians' threatened experiences—shaped by external factors and individual characteristics—and their perceived safety. Data are from ten study areas in Oregon, USA, including surveys of individuals (<em>n</em> = 551) about walking behaviors, experiences of feeling threatened, and individual characteristics, as well as data on factors related to pedestrian risk, particularly crashes, within a half-mile buffer around the respondents' addresses. We examined the relationships among these factors by estimating structural equation models. The most significant factor affecting perceived safety is pedestrians' threatened experiences due to motorists' behaviors, followed by threatened experiences due to facilities. Vehicle volume (proxy), along with the number of all types of crashes, age, gender, and walking frequency, influences perceived safety, fully mediated by threatened experiences. On the other hand, the size of parks and the presence of disabilities directly affect perceived safety, independent of prior experiences. These findings suggest that changing pedestrians' perceived safety may be challenging without addressing their experiences of being threatened by drivers' inappropriate or aggressive behaviors. Therefore, transportation planning and facility design that aim to reduce driving errors or encourage safer driving are essential for improving pedestrians' experiences, which in turn positively influence their perceived safety.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"118 ","pages":"Article 103511"},"PeriodicalIF":4.4,"publicationDate":"2026-01-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145978290","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A systematic review of gaming effects on driving related skills 游戏对驾驶相关技能影响的系统综述

IF 4.4 2区工程技术 Q1 PSYCHOLOGY, APPLIED

Transportation Research Part F-Traffic Psychology and Behaviour

Pub Date : 2026-01-07 DOI: 10.1016/j.trf.2026.103510

Rathna Bharathi Seetharaman , Peter M. Allen , Helen Keyes , David G. Pearson

Background

Road traffic accidents (RTAs) rank as the twelfth leading cause of death globally. While several factors contribute to RTAs, diminished visual and cognitive abilities remain an often-overlooked cause, negatively affecting driving performance. The “learning to learn” hypothesis suggests that playing games enhances general learning abilities and executive control mechanisms, enabling to transfer acquired-skills across real-world tasks. Considering this potential, this review investigates existing research on the connection between playing different types of games and driving behaviour.

Methods

Following PRISMA guidelines, a narrative synthesis was conducted with an effect size analysis using R. Studies were categorised based on game type, intervention duration, and outcome measures related to driving skills.

Results

The review analysed the findings from three interventional and nine observational studies. Effect size analysis of observational studies revealed a positive association between gaming experience on computerised driving tasks (g = 0.96, 95 % CI: 0.63, 1.28). Interventional studies suggested that driving racing games, when played for 8–10 h in total, can improve short-term performance in computerised and on-road driving tasks.

Conclusions

Available evidence suggests a significant positive association between gaming and computerised driving task outcomes. Furthermore, driving-specific gaming interventions have a significant effect on simulator tasks and on-road skills. However, to establish gaming interventions, further research is needed to analyse the effect of different gaming genres on different skills that are necessary for driving.

Application

Standardising interventional methodologies and driving variables are essential for providing reliable evidence. Developing evidence-based gaming interventions requires well-defined protocols and game selection criteria.

道路交通事故是全球第12大死亡原因。虽然有几个因素导致rta，但视觉和认知能力下降仍然是一个经常被忽视的原因，对驾驶性能产生负面影响。“学会学习”假说认为，玩游戏可以提高一般的学习能力和执行控制机制，从而使获得的技能能够在现实世界的任务中转移。考虑到这种可能性，本文将调查现有的关于玩不同类型游戏与驾驶行为之间联系的研究。方法遵循PRISMA指南，使用r进行叙事综合和效应量分析。研究根据游戏类型、干预持续时间和与驾驶技能相关的结果测量进行分类。结果本综述分析了3项干预性研究和9项观察性研究的结果。观察性研究的效应大小分析显示，游戏体验与电脑驾驶任务之间存在正相关关系（g = 0.96, 95% CI: 0.63, 1.28）。干预性研究表明，驾驶赛车游戏，如果总共玩8-10小时，可以提高电脑和公路驾驶任务的短期表现。现有证据表明，游戏与电脑驾驶任务结果之间存在显著的正相关。此外，驾驶游戏干预对模拟任务和道路技能有显著影响。然而，为了建立游戏干预，需要进一步研究分析不同游戏类型对驾驶所需的不同技能的影响。应用标准化的干预方法和驱动变量对于提供可靠的证据至关重要。开发基于证据的游戏干预需要明确的协议和游戏选择标准。

{"title":"A systematic review of gaming effects on driving related skills","authors":"Rathna Bharathi Seetharaman , Peter M. Allen , Helen Keyes , David G. Pearson","doi":"10.1016/j.trf.2026.103510","DOIUrl":"10.1016/j.trf.2026.103510","url":null,"abstract":"<div><h3>Background</h3><div>Road traffic accidents (RTAs) rank as the twelfth leading cause of death globally. While several factors contribute to RTAs, diminished visual and cognitive abilities remain an often-overlooked cause, negatively affecting driving performance. The “learning to learn” hypothesis suggests that playing games enhances general learning abilities and executive control mechanisms, enabling to transfer acquired-skills across real-world tasks. Considering this potential, this review investigates existing research on the connection between playing different types of games and driving behaviour.</div></div><div><h3>Methods</h3><div>Following PRISMA guidelines, a narrative synthesis was conducted with an effect size analysis using R. Studies were categorised based on game type, intervention duration, and outcome measures related to driving skills.</div></div><div><h3>Results</h3><div>The review analysed the findings from three interventional and nine observational studies. Effect size analysis of observational studies revealed a positive association between gaming experience on computerised driving tasks (g = 0.96, 95 % CI: 0.63, 1.28). Interventional studies suggested that driving racing games, when played for 8–10 h in total, can improve short-term performance in computerised and on-road driving tasks.</div></div><div><h3>Conclusions</h3><div>Available evidence suggests a significant positive association between gaming and computerised driving task outcomes. Furthermore, driving-specific gaming interventions have a significant effect on simulator tasks and on-road skills. However, to establish gaming interventions, further research is needed to analyse the effect of different gaming genres on different skills that are necessary for driving.</div></div><div><h3>Application</h3><div>Standardising interventional methodologies and driving variables are essential for providing reliable evidence. Developing evidence-based gaming interventions requires well-defined protocols and game selection criteria.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"118 ","pages":"Article 103510"},"PeriodicalIF":4.4,"publicationDate":"2026-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145927608","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Research on the influencing factors of human Drivers' Trust in fully autonomous vehicles 全自动驾驶汽车中人类驾驶员信任的影响因素研究

IF 4.4 2区工程技术 Q1 PSYCHOLOGY, APPLIED

Transportation Research Part F-Traffic Psychology and Behaviour

Pub Date : 2026-01-07 DOI: 10.1016/j.trf.2026.103509

Yifeng Cai , Yongkang Chen , Runting Tang , Wu Song

Fully Autonomous Vehicles (FAVs), as a key driver of future mobility, have demonstrated the potential to reduce traffic accidents. However, the factors influencing human drivers' trust in FAVs remain unclear. A critical issue lies in identifying these factors, particularly how the design features of FAVs affect trust. This study proposes a conceptual model incorporating empathy, information transparency, subjective knowledge, social influence, perceived risk, self-efficacy, and trust. A survey of 175 respondents was conducted, and the results were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). By examining how perceived risk and self-efficacy contribute to human drivers' trust in FAVs and the interrelations among these constructs, this study offers deeper insights into the psychological underpinnings of trust in fully autonomous vehicles. The findings reveal that trust is significantly influenced by empathy, information transparency, social influence, subjective knowledge, perceived risk, and self-efficacy. However, the effect of information transparency on perceived risk is not significant, nor is there a significant linear relationship between empathy and self-efficacy. Overall, this study enhances our understanding of human drivers' trust in FAVs and provides valuable guidance for policymakers and technology developers to devise targeted optimization strategies aimed at effectively calibrating such trust.

全自动驾驶汽车（fav）作为未来出行的关键驱动因素，已经证明了减少交通事故的潜力。然而，影响人类驾驶员对自动驾驶汽车信任的因素仍不清楚。关键问题在于确定这些因素，特别是fav的设计特征如何影响信任。本研究提出一个包含共情、资讯透明度、主观知识、社会影响力、感知风险、自我效能和信任的概念模型。对175名受访者进行了调查，并使用偏最小二乘结构方程模型（PLS-SEM）对结果进行了分析。通过研究感知风险和自我效能如何影响人类驾驶员对自动驾驶汽车的信任，以及这些结构之间的相互关系，本研究为完全自动驾驶汽车信任的心理基础提供了更深入的见解。研究发现，共情、信息透明度、社会影响力、主观知识、感知风险和自我效能显著影响信任。然而，信息透明度对感知风险的影响不显著，同理心与自我效能感之间也不存在显著的线性关系。总体而言，本研究增强了我们对人类驾驶员对自动驾驶汽车的信任的理解，并为政策制定者和技术开发人员设计有针对性的优化策略提供了有价值的指导，旨在有效地校准这种信任。

{"title":"Research on the influencing factors of human Drivers' Trust in fully autonomous vehicles","authors":"Yifeng Cai , Yongkang Chen , Runting Tang , Wu Song","doi":"10.1016/j.trf.2026.103509","DOIUrl":"10.1016/j.trf.2026.103509","url":null,"abstract":"<div><div>Fully Autonomous Vehicles (FAVs), as a key driver of future mobility, have demonstrated the potential to reduce traffic accidents. However, the factors influencing human drivers' trust in FAVs remain unclear. A critical issue lies in identifying these factors, particularly how the design features of FAVs affect trust. This study proposes a conceptual model incorporating empathy, information transparency, subjective knowledge, social influence, perceived risk, self-efficacy, and trust. A survey of 175 respondents was conducted, and the results were analyzed using Partial Least Squares Structural Equation Modeling (PLS-SEM). By examining how perceived risk and self-efficacy contribute to human drivers' trust in FAVs and the interrelations among these constructs, this study offers deeper insights into the psychological underpinnings of trust in fully autonomous vehicles. The findings reveal that trust is significantly influenced by empathy, information transparency, social influence, subjective knowledge, perceived risk, and self-efficacy. However, the effect of information transparency on perceived risk is not significant, nor is there a significant linear relationship between empathy and self-efficacy. Overall, this study enhances our understanding of human drivers' trust in FAVs and provides valuable guidance for policymakers and technology developers to devise targeted optimization strategies aimed at effectively calibrating such trust.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"118 ","pages":"Article 103509"},"PeriodicalIF":4.4,"publicationDate":"2026-01-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145927607","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Medical cannabis and driving in Australia: Results from the cannabis as medicine survey 2022–2023 (CAMS-22) 澳大利亚医用大麻和驾驶：大麻作为药物调查2022-2023 （CAMS-22）的结果

IF 4.4 2区工程技术 Q1 PSYCHOLOGY, APPLIED

Transportation Research Part F-Traffic Psychology and Behaviour

Pub Date : 2026-01-06 DOI: 10.1016/j.trf.2025.103466

Thomas R. Arkell , Llewellyn Mills , Jonathon C. Arnold , Anastasia Suraev , Sarah V. Abelev , Cilla Zhou , Nicholas Lintzeris , Iain S. McGregor

As access to medical cannabis continues to expand, understanding how patients perceive and respond to driving-related risks is important for road safety. We conducted a cross-sectional online survey of Australians using cannabis for a medical condition between December 2022 and April 2023. In addition to collecting demographic and clinical information, we assessed self-reported driving under the influence of cannabis (DUIC, defined here as ‘driving while high’), driving-related behaviours, and beliefs about impairment. Binary logistic regression was used to identify predictors of past-year DUIC. Of the 2,609 respondents who had driven in the past 12 months, 73 % (N = 1905) were accessing prescribed medicinal cannabis and 28.3 % (N = 750) reported DUIC. Several factors were associated with significantly increased odds of DUIC, including more frequent medical cannabis use, being male, using illicit and smoked cannabis, and believing that cannabis does not impair driving. The most common reason for DUIC was respondents thinking they were unimpaired (N = 518, 69.1 %). While 69 % (N = 1,790) reported that roadside drug testing deterred them from driving after cannabis use, 51 % (N = 1,340) also indicated it influenced their treatment decisions. These findings reaffirm trends identified in earlier CAMS studies and align with international literature demonstrating that perceived risk and enforcement significantly influence DUIC behaviour. Efforts to reduce DUIC among medical cannabis users need to account for the nuances of therapeutic use, noting that high-visibility enforcement strategies like roadside drug testing can reduce risky behaviours but may also restrict treatment choices. Policymakers must strike a balance between road safety and equitable access to medical cannabis.

随着获得医用大麻的机会不断扩大，了解患者如何感知和应对与驾驶有关的风险对道路安全至关重要。我们对2022年12月至2023年4月期间因医疗原因使用大麻的澳大利亚人进行了一项横断面在线调查。除了收集人口统计和临床信息外，我们还评估了自述的大麻影响下驾驶（DUIC，这里定义为“醉酒驾驶”）、驾驶相关行为和对损伤的信念。二元逻辑回归用于识别过去一年DUIC的预测因子。在过去12个月中驾车的2,609名受访者中，73% （N = 1905）获得了处方药用大麻，28.3% （N = 750）报告了DUIC。有几个因素与酒后驾车的几率显著增加有关，包括更频繁地使用医用大麻、男性、使用非法和吸烟大麻，以及认为大麻不会影响驾驶。DUIC最常见的原因是受访者认为自己没有受到损害（N = 518, 69.1%）。69% （N = 1790）的人报告说，路边药物测试阻止了他们在使用大麻后开车，51% （N = 1340）的人也表示，这影响了他们的治疗决定。这些发现重申了早期CAMS研究确定的趋势，并与国际文献一致，表明感知风险和执法显著影响DUIC行为。减少医用大麻使用者DUIC的努力需要考虑到治疗用途的细微差别，注意到路边药物检测等高能见度的执法战略可以减少危险行为，但也可能限制治疗选择。决策者必须在道路安全和公平获得医用大麻之间取得平衡。

{"title":"Medical cannabis and driving in Australia: Results from the cannabis as medicine survey 2022–2023 (CAMS-22)","authors":"Thomas R. Arkell , Llewellyn Mills , Jonathon C. Arnold , Anastasia Suraev , Sarah V. Abelev , Cilla Zhou , Nicholas Lintzeris , Iain S. McGregor","doi":"10.1016/j.trf.2025.103466","DOIUrl":"10.1016/j.trf.2025.103466","url":null,"abstract":"<div><div>As access to medical cannabis continues to expand, understanding how patients perceive and respond to driving-related risks is important for road safety. We conducted a cross-sectional online survey of Australians using cannabis for a medical condition between December 2022 and April 2023. In addition to collecting demographic and clinical information, we assessed self-reported driving under the influence of cannabis (DUIC, defined here as ‘driving while high’), driving-related behaviours, and beliefs about impairment. Binary logistic regression was used to identify predictors of past-year DUIC. Of the 2,609 respondents who had driven in the past 12 months, 73 % (N = 1905) were accessing prescribed medicinal cannabis and 28.3 % (N = 750) reported DUIC. Several factors were associated with significantly increased odds of DUIC, including more frequent medical cannabis use, being male, using illicit and smoked cannabis, and believing that cannabis does not impair driving. The most common reason for DUIC was respondents thinking they were unimpaired (N = 518, 69.1 %). While 69 % (N = 1,790) reported that roadside drug testing deterred them from driving after cannabis use, 51 % (N = 1,340) also indicated it influenced their treatment decisions. These findings reaffirm trends identified in earlier CAMS studies and align with international literature demonstrating that perceived risk and enforcement significantly influence DUIC behaviour. Efforts to reduce DUIC among medical cannabis users need to account for the nuances of therapeutic use, noting that high-visibility enforcement strategies like roadside drug testing can reduce risky behaviours but may also restrict treatment choices. Policymakers must strike a balance between road safety and equitable access to medical cannabis.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"118 ","pages":"Article 103466"},"PeriodicalIF":4.4,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145927489","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

When driving becomes enjoyable: the role of hedonic motivation and interaction quality in the adoption of autonomous vehicles 当驾驶变得愉快：享乐动机和互动质量在采用自动驾驶汽车中的作用

IF 4.4 2区工程技术 Q1 PSYCHOLOGY, APPLIED

Transportation Research Part F-Traffic Psychology and Behaviour

Pub Date : 2026-01-06 DOI: 10.1016/j.trf.2026.103507

Dong Liu , Taotao Liu , Sangbum Son , Muwen Wang

As autonomous vehicle (AV) technologies continue to advance, users' adoption intentions have become a critical factor in promoting the societal diffusion and commercial success of AVs. This study integrates the Theory of Planned Behavior (TPB) with affective motivation theory to develop a comprehensive framework encompassing affect, interaction, cognition, and behavior. It systematically examines how hedonic motivation and perceived interaction quality influence the three core TPB components (attitude, subjective norm, and perceived behavioral control) and subsequently shape behavioral intention toward AV use. A video-based scenario experiment and a two-stage survey yielded 428 valid responses. Partial least squares structural equation modeling (PLS-SEM) was employed to test the hypothesized relationships and mediation effects. The results show that hedonic motivation significantly enhances perceived interaction quality. In turn, interaction quality positively affects user attitude and perceived behavioral control, with a marginal influence on subjective norm. The impact of hedonic motivation on behavioral intention is fully mediated through the sequential path of interaction quality and TPB-related cognitions. Multi-group analysis further reveals that both users' hedonic orientation and perceived interaction level significantly moderate the structural pathways. Theoretically, this research extends the emotional dimension of TPB by highlighting the mediating role of interaction quality in the adoption mechanism of AVs. Practically, the findings offer empirical support and actionable insights for the design of affect-aware human–machine interfaces (HMIs), optimization of user experience, and segmentation strategies based on affective preferences in intelligent vehicle systems.

随着自动驾驶汽车（AV）技术的不断进步，用户的采用意愿已成为推动自动驾驶汽车社会普及和商业成功的关键因素。本研究将计划行为理论与情感动机理论相结合，建立了一个涵盖情感、互动、认知和行为的综合框架。本研究系统地考察了享乐动机和感知互动质量如何影响三个核心TPB组成部分（态度、主观规范和感知行为控制），并随后形成AV使用的行为意向。基于视频的场景实验和两阶段调查共获得428份有效回复。采用偏最小二乘结构方程模型（PLS-SEM）对假设关系和中介效应进行检验。结果表明，享乐动机显著提高了感知互动质量。交互质量对用户态度和感知行为控制有正向影响，对主观规范影响不大。快乐动机对行为意向的影响完全通过互动质量和tbp相关认知的顺序路径介导。多群体分析进一步表明，用户的享乐取向和感知交互水平显著调节了结构通路。从理论上讲，本研究通过突出互动质量在自动驾驶汽车采用机制中的中介作用，拓展了TPB的情感维度。研究结果为智能汽车系统中情感感知人机界面（hmi）的设计、用户体验的优化以及基于情感偏好的细分策略提供了实证支持和可操作的见解。

{"title":"When driving becomes enjoyable: the role of hedonic motivation and interaction quality in the adoption of autonomous vehicles","authors":"Dong Liu , Taotao Liu , Sangbum Son , Muwen Wang","doi":"10.1016/j.trf.2026.103507","DOIUrl":"10.1016/j.trf.2026.103507","url":null,"abstract":"<div><div>As autonomous vehicle (AV) technologies continue to advance, users' adoption intentions have become a critical factor in promoting the societal diffusion and commercial success of AVs. This study integrates the Theory of Planned Behavior (TPB) with affective motivation theory to develop a comprehensive framework encompassing affect, interaction, cognition, and behavior. It systematically examines how hedonic motivation and perceived interaction quality influence the three core TPB components (attitude, subjective norm, and perceived behavioral control) and subsequently shape behavioral intention toward AV use. A video-based scenario experiment and a two-stage survey yielded 428 valid responses. Partial least squares structural equation modeling (PLS-SEM) was employed to test the hypothesized relationships and mediation effects. The results show that hedonic motivation significantly enhances perceived interaction quality. In turn, interaction quality positively affects user attitude and perceived behavioral control, with a marginal influence on subjective norm. The impact of hedonic motivation on behavioral intention is fully mediated through the sequential path of interaction quality and TPB-related cognitions. Multi-group analysis further reveals that both users' hedonic orientation and perceived interaction level significantly moderate the structural pathways. Theoretically, this research extends the emotional dimension of TPB by highlighting the mediating role of interaction quality in the adoption mechanism of AVs. Practically, the findings offer empirical support and actionable insights for the design of affect-aware human–machine interfaces (HMIs), optimization of user experience, and segmentation strategies based on affective preferences in intelligent vehicle systems.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"118 ","pages":"Article 103507"},"PeriodicalIF":4.4,"publicationDate":"2026-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145927606","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Exploring user heterogeneity in autonomous vehicle acceptance: A moderated mediation model based on extended TAM 探索自动驾驶汽车接受度中的用户异质性：一个基于扩展TAM的调节中介模型

IF 4.4 2区工程技术 Q1 PSYCHOLOGY, APPLIED

Transportation Research Part F-Traffic Psychology and Behaviour

Pub Date : 2026-01-05 DOI: 10.1016/j.trf.2025.103502

Ulugbek Vahobjon Ugli Ismatullaev , Sang-Ho Kim , Kyuho Maeng

Autonomous vehicles (AVs) have the potential to revolutionize personal mobility. However, public acceptance remains a significant barrier to their adoption. This study developed an extended technology acceptance model by integrating behavioral, technological, and personality-related factors. Using survey data from 509 respondents in South Korea, we applied partial least squares structural equation modeling and bootstrapped multi-group analysis to examine moderated mediation effects across five demographic variables: age, gender, education, income, and driving experience. The results indicate that intention to use is the strongest predictor of actual use, with attitude and trust exerting significant influences on behavioral intentions. Technological perceptions, such as relative advantage, compatibility, and complexity, also influenced perceived usefulness, ease of use, and trust. Personality traits played a minimal direct role: openness was positively associated with subjective norms, while neuroticism was linked to privacy concerns, but other hypothesized effects were not statistically supported. Significant subgroup differences emerged, with gender and income being the most influential moderators of indirect pathways, highlighting the conditional nature of AV adoption. These findings extend the TAM framework by integrating demographic moderators and personality factors while underscoring their limited explanatory contribution, and they provide practical implications for tailoring AV deployment strategies to diverse user groups.

自动驾驶汽车（AVs）有可能彻底改变个人出行方式。然而，公众接受度仍然是采用它们的一个重大障碍。本研究通过整合行为、技术和个性相关因素，建立了一个扩展的技术接受模型。利用509名韩国受访者的调查数据，我们应用偏最小二乘结构方程模型和自引导多组分析来检验年龄、性别、教育程度、收入和驾驶经验这五个人口变量的调节中介效应。结果表明，使用意向是实际使用的最强预测因子，态度和信任对行为意向有显著影响。技术感知，如相对优势、兼容性和复杂性，也会影响感知的有用性、易用性和信任。性格特征的直接作用最小：开放性与主观规范呈正相关，而神经质与隐私担忧相关，但其他假设的影响没有统计支持。出现了显著的亚组差异，性别和收入是间接途径最具影响力的调节因素，突出了AV采用的条件性质。这些发现通过整合人口统计调节因子和个性因素扩展了TAM框架，同时强调了它们有限的解释贡献，并为针对不同用户群体定制AV部署策略提供了实际意义。

{"title":"Exploring user heterogeneity in autonomous vehicle acceptance: A moderated mediation model based on extended TAM","authors":"Ulugbek Vahobjon Ugli Ismatullaev , Sang-Ho Kim , Kyuho Maeng","doi":"10.1016/j.trf.2025.103502","DOIUrl":"10.1016/j.trf.2025.103502","url":null,"abstract":"<div><div>Autonomous vehicles (AVs) have the potential to revolutionize personal mobility. However, public acceptance remains a significant barrier to their adoption. This study developed an extended technology acceptance model by integrating behavioral, technological, and personality-related factors. Using survey data from 509 respondents in South Korea, we applied partial least squares structural equation modeling and bootstrapped multi-group analysis to examine moderated mediation effects across five demographic variables: age, gender, education, income, and driving experience. The results indicate that intention to use is the strongest predictor of actual use, with attitude and trust exerting significant influences on behavioral intentions. Technological perceptions, such as relative advantage, compatibility, and complexity, also influenced perceived usefulness, ease of use, and trust. Personality traits played a minimal direct role: openness was positively associated with subjective norms, while neuroticism was linked to privacy concerns, but other hypothesized effects were not statistically supported. Significant subgroup differences emerged, with gender and income being the most influential moderators of indirect pathways, highlighting the conditional nature of AV adoption. These findings extend the TAM framework by integrating demographic moderators and personality factors while underscoring their limited explanatory contribution, and they provide practical implications for tailoring AV deployment strategies to diverse user groups.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"118 ","pages":"Article 103502"},"PeriodicalIF":4.4,"publicationDate":"2026-01-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145927609","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Modelling young distracted driver's collision avoidance behaviour in a connected environment 模拟年轻分心驾驶员在互联环境中的避碰行为

IF 4.4 2区工程技术 Q1 PSYCHOLOGY, APPLIED

Transportation Research Part F-Traffic Psychology and Behaviour

Pub Date : 2025-12-31 DOI: 10.1016/j.trf.2025.103495

Qikai Qu , Yasir Ali , Yongjun Shen , Qiong Bao , Md Mazharul Haque

A connected environment can assist drivers in collision-prone situations, thereby minimising consequent crash risk. However, whether the same connected environment can also assist distracted drivers remains unclear. As such, this study examines the influence of a connected environment on young distracted drivers' collision avoidance behaviour. Fifty-one drivers with valid driving licenses participated in the connected environment driving simulator experiment. To understand the effects of the connected environment on warning hazard response time and warning collision avoidance time, separate grouped random parameters hazard-based duration models with heterogeneity in means were developed. The results suggest that relative to no warning information, young distracted drivers exhibited a faster hazard response and better collision avoidance performance in the connected environment. When specific warning information was provided, the hazard response and collision avoidance performance of young distracted drivers were better than simple warning information. When early warning information was provided, young distracted drivers demonstrated enhanced hazard response and collision avoidance performance in comparison to the late warnings. Further, the timing of warning information significantly affected drivers' hazard response, while the informational content markedly influenced their collision avoidance performance. These findings can assist in the development of advanced collision avoidance warning systems in a connected environment.

连接的环境可以在容易发生碰撞的情况下帮助驾驶员，从而最大限度地降低随之而来的碰撞风险。然而，同样的联网环境是否也能帮助分心的司机仍不清楚。因此，本研究考察了互联环境对年轻分心司机避碰行为的影响。51名持有有效驾照的驾驶员参加了互联环境驾驶模拟器实验。为了解互联环境对预警危险响应时间和预警避碰时间的影响，建立了均值异质性的独立分组随机参数风险持续时间模型。结果表明，相对于没有警告信息，年轻分心司机在互联环境中表现出更快的危险反应和更好的碰撞避免性能。当提供特定的警告信息时，年轻分心驾驶员的危险响应和避碰性能优于简单的警告信息。当提供早期预警信息时，与晚期预警相比，分心的年轻司机表现出更强的危险反应和碰撞避免性能。此外，警示信息的时间对驾驶员的危险反应有显著影响，而警示信息的内容对驾驶员的避碰行为有显著影响。这些发现有助于在互联环境中开发先进的避碰预警系统。

{"title":"Modelling young distracted driver's collision avoidance behaviour in a connected environment","authors":"Qikai Qu , Yasir Ali , Yongjun Shen , Qiong Bao , Md Mazharul Haque","doi":"10.1016/j.trf.2025.103495","DOIUrl":"10.1016/j.trf.2025.103495","url":null,"abstract":"<div><div>A connected environment can assist drivers in collision-prone situations, thereby minimising consequent crash risk. However, whether the same connected environment can also assist distracted drivers remains unclear. As such, this study examines the influence of a connected environment on young distracted drivers' collision avoidance behaviour. Fifty-one drivers with valid driving licenses participated in the connected environment driving simulator experiment. To understand the effects of the connected environment on warning hazard response time and warning collision avoidance time, separate grouped random parameters hazard-based duration models with heterogeneity in means were developed. The results suggest that relative to no warning information, young distracted drivers exhibited a faster hazard response and better collision avoidance performance in the connected environment. When specific warning information was provided, the hazard response and collision avoidance performance of young distracted drivers were better than simple warning information. When early warning information was provided, young distracted drivers demonstrated enhanced hazard response and collision avoidance performance in comparison to the late warnings. Further, the timing of warning information significantly affected drivers' hazard response, while the informational content markedly influenced their collision avoidance performance. These findings can assist in the development of advanced collision avoidance warning systems in a connected environment.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"118 ","pages":"Article 103495"},"PeriodicalIF":4.4,"publicationDate":"2025-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145885741","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Life satisfaction, risky and aggressive driving, and crash involvement: Evidence within a contextual-mediated framework 生活满意度，危险和攻击性驾驶，和碰撞参与：证据在一个情境介导的框架

IF 4.4 2区工程技术 Q1 PSYCHOLOGY, APPLIED

Transportation Research Part F-Traffic Psychology and Behaviour

Pub Date : 2025-12-30 DOI: 10.1016/j.trf.2025.103496

Jorge-Eduardo Martínez-Pérez , Eduardo Martínez-Gabaldón , Jose-María Abellán-Perpiñán , Fernando-Ignacio Sánchez-Martínez

Road traffic crashes remain a critical public-health issue, with human factors implicated in the vast majority of incidents. This study applies the contextual-mediated model of Sümer to test whether life satisfaction, as a distal factor, relates to self-reported crash involvement indirectly through proximal behaviours (risky and aggressive driving). The survey responses of 4077 Spanish adults are analysed. Life satisfaction is assessed with the Satisfaction with Life Scale; risky driving captured any alcohol/drug-impaired driving (or riding with an impaired driver); aggressive driving was measured on a standardised four-item scale. Logistic and linear models adjusted for age, gender, exposure, education, marital status, and region show that higher life satisfaction is associated with lower risky driving (OR ≈ 0.70) and lower aggressive driving (≈ −0.14 SD). In crash models, the life-satisfaction coefficient is negative, and it is attenuated once proximal behaviours are added in the main specification. Moreover, Karlson-Holm-Breen decomposition indicates a substantive indirect component (Diff ≈ −0.054 in log-odds), with risky driving accounting for the larger share of mediation. Robustness checks using a median split of life satisfaction (Diff ≈ −0.081) and a licensed-driver subsample (Diff ≈ −0.066) show similar patterns, with the direct association becoming marginal/non-significant after including proximal factors. Findings highlight behavioural pathways and suggest complementing road-safety programmes targeting anger regulation and risky styles with initiatives that foster evaluative well-being to curb risky behaviours and reduce crashes in Spain.

道路交通事故仍然是一个严重的公共卫生问题，绝大多数事故都与人为因素有关。本研究采用情景介导的 mer模型来测试生活满意度作为一个远端因素，是否通过近端行为（危险和攻击性驾驶）间接与自我报告的碰撞卷入相关。对4077名西班牙成年人的调查结果进行了分析。用生活满意度量表评估生活满意度；危险驾驶记录任何酒后/吸毒后驾驶（或与醉酒司机一起驾驶）；攻击性驾驶是用一个标准化的四项量表来衡量的。调整了年龄、性别、暴露、教育、婚姻状况和地区的Logistic和线性模型表明，较高的生活满意度与较低的危险驾驶（OR≈0.70）和较低的攻击性驾驶（SD≈−0.14）相关。在碰撞模型中，生活满意度系数为负，一旦在主规范中加入近端行为，它就会减弱。此外，Karlson-Holm-Breen分解表明存在实质性的间接成分（对数赔率中的Diff≈−0.054），其中风险驾驶占更大的中介份额。使用生活满意度中位数分裂（Diff≈−0.081）和有执照的司机子样本（Diff≈−0.066）进行稳健性检查显示出类似的模式，在包括近端因素后，直接关联变得边缘/不显著。研究结果强调了行为途径，并建议将针对愤怒调节和冒险风格的道路安全计划与促进评估福利的举措相辅相成，以遏制危险行为并减少西班牙的撞车事故。

{"title":"Life satisfaction, risky and aggressive driving, and crash involvement: Evidence within a contextual-mediated framework","authors":"Jorge-Eduardo Martínez-Pérez , Eduardo Martínez-Gabaldón , Jose-María Abellán-Perpiñán , Fernando-Ignacio Sánchez-Martínez","doi":"10.1016/j.trf.2025.103496","DOIUrl":"10.1016/j.trf.2025.103496","url":null,"abstract":"<div><div>Road traffic crashes remain a critical public-health issue, with human factors implicated in the vast majority of incidents. This study applies the contextual-mediated model of Sümer to test whether life satisfaction, as a distal factor, relates to self-reported crash involvement indirectly through proximal behaviours (risky and aggressive driving). The survey responses of 4077 Spanish adults are analysed. Life satisfaction is assessed with the Satisfaction with Life Scale; risky driving captured any alcohol/drug-impaired driving (or riding with an impaired driver); aggressive driving was measured on a standardised four-item scale. Logistic and linear models adjusted for age, gender, exposure, education, marital status, and region show that higher life satisfaction is associated with lower risky driving (OR ≈ 0.70) and lower aggressive driving (≈ −0.14 SD). In crash models, the life-satisfaction coefficient is negative, and it is attenuated once proximal behaviours are added in the main specification. Moreover, Karlson-Holm-Breen decomposition indicates a substantive indirect component (Diff ≈ −0.054 in log-odds), with risky driving accounting for the larger share of mediation. Robustness checks using a median split of life satisfaction (Diff ≈ −0.081) and a licensed-driver subsample (Diff ≈ −0.066) show similar patterns, with the direct association becoming marginal/non-significant after including proximal factors. Findings highlight behavioural pathways and suggest complementing road-safety programmes targeting anger regulation and risky styles with initiatives that foster evaluative well-being to curb risky behaviours and reduce crashes in Spain.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"118 ","pages":"Article 103496"},"PeriodicalIF":4.4,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145885745","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

AI vs. Humans: Comparing road user intention recognition performance 人工智能与人类：比较道路使用者意图识别性能

IF 4.4 2区工程技术 Q1 PSYCHOLOGY, APPLIED

Transportation Research Part F-Traffic Psychology and Behaviour

Pub Date : 2025-12-30 DOI: 10.1016/j.trf.2025.103491

Koen Vellenga , H. Joe Steinhauer , Göran Falkman , Jonas Andersson , Anders Sjögren

Anticipating the behavior of other road users is critical for safe driving. To anticipate the behavior of other road users in a timely manner, it is essential to recognize their intentions. Although artificial intelligence (AI)-based intention recognition systems for traffic scenarios have advanced significantly, their performance relative to human road user intention recognition (RUIR) remains largely unexplored. To address this gap, we conducted an experiment comparing the RUIR performance of human participants and a state-of-the-art end-to-end video recognition AI model on a set of 25 video scenarios. The selected scenarios offered a balanced representation of various road user types and a range of intention maneuvers. Among human participants (N=161), we found no statistically significant differences in RUIR performance with respect to age, self-perceived driving skill, annual driven kilometers, or years of driving experience. However, the average human participant exhibited slightly lower RUIR performance than the AI models.

预测其他道路使用者的行为对安全驾驶至关重要。为了及时预测其他道路使用者的行为，认识他们的意图是至关重要的。尽管基于人工智能（AI）的交通场景意图识别系统已经取得了显着进步，但它们相对于人类道路使用者意图识别（RUIR）的性能在很大程度上仍未得到探索。为了解决这一差距，我们进行了一项实验，比较了人类参与者和最先进的端到端视频识别AI模型在25个视频场景中的RUIR表现。所选择的场景提供了各种道路使用者类型和一系列意图操作的平衡代表。在人类参与者（N=161）中，我们发现在年龄、自认为的驾驶技能、年驾驶公里数或驾驶经验方面，RUIR表现没有统计学上的显著差异。然而，人类参与者的平均RUIR表现略低于人工智能模型。

{"title":"AI vs. Humans: Comparing road user intention recognition performance","authors":"Koen Vellenga , H. Joe Steinhauer , Göran Falkman , Jonas Andersson , Anders Sjögren","doi":"10.1016/j.trf.2025.103491","DOIUrl":"10.1016/j.trf.2025.103491","url":null,"abstract":"<div><div>Anticipating the behavior of other road users is critical for safe driving. To anticipate the behavior of other road users in a timely manner, it is essential to recognize their intentions. Although artificial intelligence (AI)-based intention recognition systems for traffic scenarios have advanced significantly, their performance relative to human road user intention recognition (RUIR) remains largely unexplored. To address this gap, we conducted an experiment comparing the RUIR performance of human participants and a state-of-the-art end-to-end video recognition AI model on a set of 25 video scenarios. The selected scenarios offered a balanced representation of various road user types and a range of intention maneuvers. Among human participants (N=161), we found no statistically significant differences in RUIR performance with respect to age, self-perceived driving skill, annual driven kilometers, or years of driving experience. However, the average human participant exhibited slightly lower RUIR performance than the AI models.</div></div>","PeriodicalId":48355,"journal":{"name":"Transportation Research Part F-Traffic Psychology and Behaviour","volume":"118 ","pages":"Article 103491"},"PeriodicalIF":4.4,"publicationDate":"2025-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145885744","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0