Driver models are essential for virtual safety assessments of automated vehicles. This study evaluates the realism of the i4Driving model, a behavioural driver model designed to emulate human-like driving, using an interactive driving Turing test combined with a think-aloud protocol in connected driving simulators. Thirty participants interacted with either a human-controlled vehicle or the i4Driving model across three motorway scenarios and rated realism, predictability, safety, and aggressiveness. Results showed that the i4Driving model was perceived as less realistic and predictable than human drivers (p < 0.001), yet participants could not reliably distinguish between the two (classification accuracy = 0.673). Think-aloud analysis revealed specific shortcomings, such as unrealistic merging and speed adaptation, alongside instances of naturalistic behaviour. These findings highlight the need for improvements in tactical decision-making and demonstrate the value of combining subjective ratings with qualitative insights for refining driver models.

Simulation is a valuable tool for traffic planning that requires reliable modeling of traffic dynamics. Free flow speeds in bicycle traffic depend on the characteristics and preferences of the bicyclists, infrastructure design, and environmental conditions. However, existing models are limited in capturing changes in the speed during free riding, thereby reducing their applicability in bicycle traffic analysis. This study advances microscopic bicycle traffic simulation by developing and evaluating simulation models for free riding dynamics, aiming to capture the heterogeneous and context-dependent effects of infrastructure design (slopes, curves, and presence of intersections) and wind on bicyclist behavior. We implement three models within SUMO —(1) context-based speed distributions, (2) a speed regression model, and (3) a physics-based speed model derived from power output— and benchmark them against built-in models. All models are evaluated using empirical trajectory data from 57 bicycle commuters in semi-controlled experiments in Sweden and Germany. Results demonstrate that the proposed models outperform the existing baselines in replicating speed patterns. In this regard, the physics-based model provides the closest alignment to observed speeds. Omitting context dependency in free riding can result in substantially larger errors in speeds on uphills and downhills, and in deceleration on curves or when crossing intersections. Context-sensitive models enhance the accuracy of bicycle traffic simulation, thereby increasing their usefulness in planning and evaluating bicycling facilities that accommodate the diverse preferences of bicyclists.

Overtaking behavior on oncoming separated rural roads with intermittent overtaking sections, so called 2+1 roads, is unique as there are predefined sections where overtaking can take place. There is a growing interest in questions related to the design of such roads, including the length and placement of overtaking lanes to improve both safety and traffic efficiency. This study applies a computer vision-based vehicle re-identification methodology to gather evidence about vehicle speeds and overtaking and merging behavior along a 2-lane segment of a 2 + 1 road on the E18 route north of Stockholm. The results show that the vehicle re-identification method performs satisfactorily with only a few missed or false detections and more than 85% of all vehicles were successfully re-identified at all the cameras. The missed detections were commonly due to the vehicle not being detected at the first sensor because of occlusion from other vehicles. The method enables new possibilities to study number of overtakings for different sub-segments, how many vehicles that a single vehicle overtake, speeds or overtaking and overtaken vehicles, space mean speed, etc. For example, a comparison of the segment speed over the entire two-lane stretch with cross-section speed measurements from the different measurement locations shows the possibility to identify the location where cross-section measurements on two-lane stretches should be placed in order to best estimate the speed on the entire two-lane stretch. The current study also provides insights into how high the speed of overtaking vehicles is and how it varies between vehicles and segments as well as how many overtakes are conducted on the merging segment. One important observation is that between 5 – 8% of the overtaking vehicles tend to try to maximize the number of overtakes on the two-lane stretch by increasing the speed up to 120–130 km/h and a direct relationship can be observed between the number of overtakings performed by a vehicle with the speed of the overtaking vehicle. 

Mätning av trafikföring har traditionellt genomförts genom punktmätningar (slang, induktiv slinga, radar) eller floating-car mätningar (GPS-utrustade fordon). Punktmätningar ger information om alla fordon som passerar en specifik punkt eller tvärsnitt på vägen medan floating carmätningar ger information om enskilda fordons körförlopp längs med en vägsträcka. Ingen av dessa mätmetoder ger således en fullständig bild av hur trafikföringen ser ut i rum och tid. På senare tid har video-baserade mätningar utvecklats och använts för att extrahera körförlopp för alla fordon som videomätningen omfattar. Detta fungerar relativt bra i stadstrafik där sträckorna som ska studeras är relativt korta. Hur stort område som kan studeras beror på kamerans vertikala avstånd till vägen, men även med drönare så är den längsta sträcka som kan studeras i dagsläget som mest upp mot 500 meter. För mätning av trafikföring på exempelvis mötesfria vägar (så kallade 2+1 vägar) behöver längre sträckor täckas för att möjliggöra analys av antal genomförda omkörningar, hastighetsvariationer över delsträckor och i samband med omkörningar etc. För att möjliggöra sådana analyser har en pilotmätning genomförts där sex stycken kameror placerats ut längs med en tvåfältssträcka på en 2+1 väg. Varje kamera fångar alla fordons körförlopp under en begränsad sträcka (ca 50 meter) men genom utveckling av en metodik för återigenkänning kopplas fordonens passager vid olika kameror samman vilket möjliggör analys av fordonens restid/genomsnittshastighet mellan kameror, om omkörningar har genomförts, etc. Den nya metodiken ger således inte bara möjlighet att studera de sträckor som kamerorna täcker utan även möjlighet att extrahera information om vad som hänt mellan kameraplaceringarna.  

The behavior of bicyclists is influenced by infrastructure, weather, other road users, and individual capabilities and preferences. The fundamentally human-powered nature of bicycling indicates that physical capabilities and perceptions of effort may also play a key role in how bicyclists ride, reflecting diverse capabilities and individual choices. Bicycling research has predominantly focused on speed and acceleration, often overlooking detailed behaviors related to perceived effort. The purpose of this study is to examine free-riding behavior --uninfluenced by other road users -- and identify correlations with contextual features such as infrastructure design (horizontal and vertical alignment) and wind speed. We introduce a method using instrumented bicycles in a semi-controlled experiment to collect detailed data on free-riding behavior, including speed, power output, and heart rate. Participants use their own bicycles equipped with sensors on designated routes and complete a survey to assess their experience and perceived effort. Experiments conducted in Germany and Sweden allow for comparative analysis of different infrastructure designs.

Många förare har ett avsökningsbeteende som systematiskt missar vissa potentiella konflikter, vilket indikerar att deras mentala modeller är felaktiga. Exempelvis förekommer att förare inte tittar efter cyklister som kommer bakifrån innan de svänger höger. Vi undersöker därför hur avsökningsbeteende och anpassning till situationen varierar med omgivande trafik, körriktning, trafikregler och icke-körrelaterade uppgifter, för att förbättra förståelsen av varför förare gör systematiska fel då de söker av omgivningen.

En körsimulatorstudie i stadsmiljö med 40 förare har genomförts. Förarna valdes baserat på självskattad körstil (offensiv/defensiv) samt vana av att cykla i stadsmiljö (erfaren/oerfaren cyklist). Deltagarna exponerades flertalet gånger för tre situationer: sväng i korsning med interagerande cyklist, övergångställe med fotgängare, samt interaktion med andra motorfordon. Under försöket har data om hur förarna kör i förhållande till omgivningen och andra trafikanter, förarnas blickbeteende, förarnas arbetsbelastning, samt hur förarna interagerar med en icke-körrelaterad uppgift samlats in. Analyserna pågår i skrivande stund. Resultaten är därför avgränsade till situationen "korsning med interagerande cyklist".

Förarna stannade eller krypkörde innan de körde ut i korsningen i 254 av de 348 passagerna med högersväng och i 331 av 348 passager med vänstersväng. I många fall sökte försökspersonerna efter informationen för tidigt, t ex tittade försökspersonerna höger/vänster i korsningen när sikten fortfarande blockerades av kringliggande byggnader i 20% av passagerna. Beroende på kategori tittar mellan 50–80% av förarna i speglarna och/eller över axeln efter cyklister innan de svänger ut i korsningen. Den kategori som tittar oftast är de defensiva förarna som även är vana cyklister.

Högersvängsolyckor är ett vanligt scenario där cyklister blir påkörda. En trolig orsak är att förare inte söker av sina döda vinklar. Frågan är varför förare glömmer, eller inte ens känner till, att de behöver titta efter cyklister som kommer bakifrån. Våra preliminära resultat ger inga tydliga svar, mer än att bilister som även har erfarenhet av att cykla i stadstrafik är bättre på att samspela med cyklister.

Previous research has shown that drivers often fail to yield to cyclists who have the right of way, particularly when turning right. This finding was confirmed in a driving simulator study. To determine whether this is a systematic issue related to the type of attention required by road users, a systematic investigation was conducted.

When travelling through an intersection, different trajectories intersect. Various rules determine who has priority, resulting in different attention targets that road users must monitor to proceed safely and in accordance with the rules. For example, when going straight on a main road, the only attentional requirements on the driver are to monitor to the left and right for pedestrians who might want to cross on the zebra crossing. On a subordinate road in the same situation, the driver also needs to check for cross traffic from the left and from the right.  

The attention targets at intersections can be classified as either ‘overt’ (visible via signs, road markings, etc.) or ‘covert’ (unmarked and have to be remembered, e.g., the need to check ‘over the right shoulder’ for cyclists and pedestrians when a driver is turning right from a main road). Additionally, they can be categorised as ‘same-mass’ (for partners in an encounter) or ‘lower-mass’ (as in the turning-right example where the cyclist has lower mass), and as ‘required’ (when one must yield to superior traffic streams) or ‘not required’ (when one has priority).   

Conflicts arise when motor vehicles cross cycle paths during turns, often due to drivers neglecting over-the-shoulder glances. This paper examines the visual attention of car drivers at urban unsignalized intersections, considering turning direction, cross traffic, and cyclist approaching from behind.

A study with 44 participants was conducted using a fixed-base driving simulator equipped with an eXtended Reality (XR)-based visual system, which provided 360° immersion and enabled eye tracking for head movements and over-the-shoulder glances. Drivers were recruited based on urban cycling experience (inexperienced/experienced) and driving style (cautious/assertive). 

Results showed that in 47.8 % of intersection approaches, participants failed to adequately check for cyclists from behind. Driver characteristics did not consistently reveal traits associated with this neglect. Although cautious drivers with urban cycling experience made fewer mistakes, all groups were generally poor at checking for cyclists from behind. Post-drive questionnaire results on rule knowledge showed that, across participants, only half of the requirements to check for cyclists approaching from behind were indicated correctly. 

Drivers are more consistent in checking for cross-traffic, suggesting the neglect to check for cyclists is systemic rather than individual. The lack of awareness among drivers about their obligation to check for cyclists can be attributed to the less obvious nature of this traffic stream, as it comes from behind and lacks signs or warnings. Checking over the shoulder is physically more effortful, and cyclists may often pre-empt collisions, reinforcing drivers’ mental models to turn without checking. To improve safety, changes counteracting systemic biases against cyclists are needed.  

Studies of bicycle traffic characteristics are essential for understanding bicyclist behavior and developing suitable microscopic models for traffic simulation. Based on empirical data on real-world bicycle traffic, obtained through video-based data collection techniques, we describe characteristics of free-riding, following, and overtaking that facilitate the simulation of bidirectional bicycle traffic. In total, we analyze data from 195 605 bicyclists across seven bicycle paths in Sweden, with five located in Stockholm and two in Gothenburg. The analysis reveals multimodal distributions of speed and lateral position due to the distinct preferences of bicyclists that vary according to the type of bicycle used. Moreover, speeds are generally highest during morning rush hours at the locations under investigation. Based on wind measurements, we conclude that there is no statistically significant effect on mean free speed from wind speeds in the range of ±3 m/s. The distribution of speed differences between bicyclists at overtakings indicates a broad range of speed disparities among bicyclists, and is significantly influenced by infrastructure design. Furthermore, infrastructure design (path width, horizontal alignment, and presence of fixed objects along the edge) and traffic volumes significantly lateral positioning. Our results demonstrate the inherent heterogeneity in the characteristics of bicyclists, underscoring the need to advance modeling to incorporate these distinct characteristics into microscopic traffic simulation.  

Microscopic traffic simulation models rely on simplifications of many aspects of driving behavior to describe traffic phenomena. Some aspects omitted by these simplifications become essential to distinguish the driving behavior of automated driving systems (ADSs) and humans. We consider the perception processes as one of such omitted aspects. The aim of this study is to evaluate the effects of an imperfect perception on the traffic flow dynamics using microscopic traffic simulation. A modeling approach is used which explicitly describes the perception performance in terms of accuracy, precision, delay, and range. Simulation experiments are presented which demonstrate possible effects on car-following and lane-changing situations. Results show that traffic phenomena such as instabilities, oscillations, and moving traffic jams can be simulated, providing a broader perspective about potential risks to the traffic flow performance. The explicit modeling of the perception has direct application on the study of mixed traffic of human and automated driving in microscopic traffic simulation.