In most recent cases, motivated by the need to conduct tests in both physical and virtual test environment, a digital twin of the physical world is often created in the virtual environment to be used for testing purposes. This approach is often used in when testing automated driving systems (ADS) which are being extensively tested by companies and cities around the world. While this approach may work well with ADS, whose behaviour can be deterministic in both physical and virtual worlds, human drivers are known to exhibit different behaviour in simulation compared to real situations.

Since human drivers are still expected to be involved in future transport systems, the difference in driving behaviour needs to be understood in order to assist in verification and validation of the transport system. Therefore, in this work, we intend to investigate the differences in driving behaviour of human drivers, comparing the same manoeuvre done on the proving ground to the manoeuvre recorded in a driving simulator. 

Data collected from a proving ground will be compared with data collected from a driving simulator (under the same manoeuvre) to analyse differences in driving behaviour between the two test environments. More specifically, there are two datasets in this work: 1) data collected from AstaZero proving ground; and 2) data collected from the "Sim IV" driving simulator at VTI.

The difference between collected data will be analysed using statistical methods. The results indicating difference in driving behaviour between the two datasets will be presented. The results will be analysed and presented in the context of verification and validation of transport systems using digital twins.

This work intends to present the differences in driving behaviour of human drivers, comparing the same manoeuvre done on the proving ground to the manoeuvre recorded in a driving simulator. The result will provide some indication on the validity of using driving simulator as a virtual test environment for verification and validation of transport systems.

På VTI har tre simulatorstudier genomförts under 2024 som använder sig av denna teknik. I dessa tre studier har utrustningen använts som körsimulator, simulator för gångare och simulator för cyklister. Vid byggandet av simulatorlösningarna så provades ett antal olika arbetssätt och inställningar för att sedan landa i hur studierna genomfördes. När studierna genomfördes så tillfrågades deltagarna i slutet av studierna om hur de tyckte det gick att använda utrustningen och om hur realistisk deras upplevelse var. Dessa svar har sedan analyserats och jämförts med historiska data från tidigare simulatorstudier.

Vid användning av ”green-screen” teknik för XR så har det varit viktigt med bra ljusförhållanden. För gångare användes extra ”trackers” för att ge ett tillfredsställande rörelsebeteende. Dessa ”trackers” behöver vara i kontakt med basstationer så med väldigt rörliga gångare kan fler basstationer behövas. XR systemen kräver mycket datorprestanda så det är viktigt att ha rätt hårdvara och en del optimering kan behövas för att få tillfredsställande prestanda. Vid långvarig användning så kan det vara bra att stänga av/starta om systemet då och då. Försökspersoner som har varit gångare eller cyklister beskriver en högre realism än förare även om alla tycker att det har varit realistiskt. Det fanns heller inga indikationer på en ökad simulatorsjuka och för cyklister så rapporterades den lägsta simulatorsjukan. De flesta tyckte att det gick bra att använda ett headset och tyckte inte att det var speciellt besvärande. Att kunna se detaljer i miljön verkar ligga på samma nivå som med skärmar.

Från dessa studier så kan vi se att huvudbaserade XR/VR lösningar ger ytterligare möjligheter till simulatorstudier och kompletterar på så sätt skärmbaserade simulatorer. En fördel är att de ger en friare känsla av att röra sig och kunna titta runt samtidigt som de kan spåra var försökspersoner tittar. En nackdel är att de är prestandakrävande vilket ger en utmaning i att skapa väldigt levande miljöer. Sammanfattningsvis så kommer VTI fortsatt att använda tekniken.

Remote operation of highly automated vehicles (HAVs) may include occasional assistance from a human remote operator that is located outside the HAVs. Remote assistance typically delegates only high-level guidance tasks to the remote operators such as authorizing a driving maneuver or specifying a new driving path. As remote assistance is fairly unexplored, there are still several research challenges. These challenges were discussed by experts from academia and industry in a multidisciplinary workshop at the 2023 IEEE Intelligent Vehicles Symposium. As a result of the workshop, this paper presents a list of most pressing research questions in the following areas: human-machine interaction and human factors, design of the remote station, design of the HAVs. It also outlines a roadmap for future research on remote assistance of HAV, thereby informing interdisciplinary studies and facilitating the benefits of HAVs before full autonomy can be reached.

The study explores the impact of light conditions on driver sleepiness. In a driving simulator experiment, 20 drivers drove both during daytime in an alert condition and then later at night after being awake since early morning. Light conditions were manipulated by driving in a simulated nighttime scenario in a dark room (1 lx) versus driving in simulated daylight in a room lit with light emitting diodes combining blue light with a yellow phosphor giving a two peaked spectrum (212 lx). Both the daylight and the darkness scenarios were driven daytime and nighttime in a 2 × 2 design. Sleepiness was measured during the four 1-hour drives in terms of subjective sleepiness ratings, divided attention ability, driving performance, heart rate variability and blink behaviour. Significant differences were found in all measured sleepiness indicators between the daytime and nighttime drives, and in most indicators for time-on- task. No significant main effects were found between simulated daylight and darkness. A psychomotor vigilance test conducted before and after each drive also showed no significant effects for lighting condition. Further research, preferably using longitudinal studies in more realistic settings on real roads, is needed to determine which behaviours and which cognitive processes that are affected when driving in daylight versus darkness.

The effect of the interaction between Cooperative, Connected and Automated Mobility (CCAM) in scenarios where the ratio of conventional and intelligent vehicles is a topic of interest for researchers and the industry. There is an expected effect on road safety and traffic efficiency stemming from the coexistence between Connected Automated Vehicles (CAVs) and other non-connected, non-automated road users like disconnected pedestrians or legacy vehicles – which we group under the term Multi-modal Road Users (MRUs). These effects, either positive or negative, might not grow linearly with increased adoption rates and, furthermore, it is realistic to expect MRUs to share the road with CAVs even as outliers in the fleet. Thus, identifying and analyzing if existing and developing safety metrics apply to a near full-CCAM environment is crucial, specifically, if human factors (which affect conventional driving) continue to influence safety and efficiency when full connection and automation is expected.This pre-study explores the literature and performs a simulation-based analysis of these risks and human factors on road safety and traffic efficiency. Starting from an exploration of the literature, where groundwork exists on existing and expected risks and mitigations, we map these risks to scenarios where full CAV driving is expected. We identify that some of the humanly influenced risk factors that affect the Dynamic Driving task (DDT) are mitigated, while some other (e.g., decision making at the design stage of CAVs) might even create more heterogeneity. Then, in the simulation stage, we identify that heterogeneity is what influences safety and efficiency, and that it is not only the ratio between CAVs and MRUs that affects convergence but also the place in which heterogeneous vehicles are placed in the flow.

This literature survey explores the domain of remote driving of road vehicles within autonomous vehicles, focusing on challenges and state-of-the-art solutions related to driving feedback, latency, support control, as well as remote driving platform and real applications. The advancement towards Level-5 autonomy faces challenges, including sensor reliability and diverse scenario feasibility. Currently, remote driving is identified as vital for commercialization, however, it comes with challenges like low situational awareness, latency, and a lack of comprehensive feedback mechanisms. Solutions proposed include enhancing visual feedback, developing haptic feedback, employing prediction techniques, and use control methods to support driver. This paper reviews the existing literature on remote driving in these fields, revealing research gaps and areas for future studies. Additionally, this paper reviews the industry applications of remote driving and shows the state-of-art use cases.

In the evolving landscape of driving simulation, achieving a comprehensive representation of urban scenarios requires an understanding of driver-pedestrian interactions. In order to conduct simulator studies on driver-pedestrian interactions, a simulation framework that allows real-time interaction between the driver and pedestrian is needed. To address this need, this paper presents an approach to driver-pedestrian co-simulation. The pedestrian simulator in our work utilizes real-time motion capture, enabling accurate representation of the pedestrian’s states in real-time. Moreover, advantages and disadvantages of the presented simulation architecture are discussed along with approaches for future improvements.

One2Many, the title of this report, refers to remote operation of vehicles where an operator handles several vehicles simultaneously. This may increase efficiency and opportunities for profitability. The objective of this report is to identify essential and relevant developments of the regulatory framework, as well as business models and working conditions for support of safe and sustainable introduction of a single person’s remote operation of multiple vehicles. 

As a starting point, this report describes the taxonomy used and a state-of-the-art study. Most literature deals with technical challenges around remote operation while non-technical challenges, as well operation of multiple vehicles, are poorly covered. This report aims at addressing this gap by considering non-technical aspects for remote operation. 

Legal aspects are described and analysed, resulting in recommendations for next steps for legislation and regulation. Furthermore, business models and working environment are discussed, taking advantage of two real world use cases: goods transport (Einride trucks), and public transportation (Ride the Future automated shuttles). 

One2Many summarises regulatory considerations in a Memorandum, and additionally concludes that: 

• Research is needed regarding legal challenges for the three different modes of remote operation (remote driving, remote assistance, remote supervision) and how to address them in future regulation to best deal with safety concerns and to support remote operation. Liability issues and concerns also need to be handled. 

• The main advantage of introducing remote operating for several vehicles per operator will most likely be uptime. The employee cost is foreseen to decrease, but potential surrounding functions need to be studied in order to determine if the cost of personnel actually would go down. 

• Several working environment considerations should be further discussed, e.g. regarding what type of controls would be most effective and safe to use, and whether it matters if an operator has a background as driver of conventional vehicles.

Föreliggande rapports titel, One2Many, syftar på fjärrstyrning av fordon där en operatör hanterar flera fordon samtidigt. Detta kan öka effektiviteten och möjligheterna till lönsamhet. Projektets syfte är att identifiera väsentliga och relevanta utvecklingar av regelverket, såväl som affärsmodeller och arbetsförhållanden för att stödja säker och hållbar introduktion av en enda persons fjärrstyrning av flera fordon. 

I avsaknad av etablerad taxonomi för styrning identifieras begreppet ”Remote operation” (fjärrstyrning) och dess tre moder: remote driving, assistance och supervision (fjärrkörning, fjärrassistans, fjärrövervakning). En nulägesbekrivning (state of the art) visar att den mesta litteraturen och de flesta initiativ inom industri och forskning behandlar tekniska utmaningar kring fjärrstyrning. Icke tekniska utmaningar, liksom samtidig styrning av flera fordon, täcks i betydligt mindre grad. Rapporten adresserar denna lucka genom att beakta icke-tekniska aspekter för fjärrstyrning. 

Juridiska frågeställningar beskrivs och analyseras, och leder fram till rekommendationer för hur lagstiftning och reglering behöver utvecklas. Vidare diskuteras affärsmodeller och arbetsmiljö utifrån två fall med automatiserade fordon i verkliga miljöer: godstransport (Einride-lastbilar) och kollektivtrafik (Ride the Future automatiserade skyttlar). 

One2Many sammanfattar regulatoriska överväganden i ett memorandum och drar dessutom slutsatserna att: 

• Det är nödvändigt med forskning som behandlar de juridiska utmaningarna för de tre olika fjärrstyrningsmoderna och deras framtida reglering för att på bästa sätt hantera säkerhetsproblem och för att stödja fjärrstyrning. Ansvarsfrågor och problem måste också behandlas. 

• Den största fördelen med att införa fjärrstyrning för flera fordon per operatör kommer med största sannolikhet att vara ”up-time”. Personalkostnaden förväntas minska, men operationernas kringaktiviteter och -funktioner måste studeras för att avgöra om personalkostnaderna faktiskt skulle minska. 

• Flera arbetsmiljöaspekter bör diskuteras vidare, till exempel om vilken typ av kontroller som skulle vara mest effektiva och säkra att använda, och om det spelar någon roll om en operatör har en bakgrund som förare av konventionella fordon, eller inte.

This report presents experimental setups and findings from the REDO project, which had been conducted between December 2019 and February 2023. Five main topics are covered in this report: 1) Effects of latency and field-of-view on driving performance; 2) Remote driving feedback and control; 3) Connectivity and mobile network support for remote driving; 4) Video transmission for remote driving; and 5) Laws and regulations concerning remote driving. Contents of this report dives into technical details and findings within each topic. Nevertheless, this report does not intend to repeat all detail and results published in scientific publications, and thus this report should be seen as complementary material to the published results.

Denna rapport presenterar studier och experiment inom REDO-projektet samt resultat ifrån dessa. Studierna är utförda mellan december 2019 och februari 2023 och rapporten täcker fem huvudämnen: 1) den effektlatens och siktvinkel har på körprestanda; 2) feedback och kontroll vid fjärrkörning; 3) uppkoppling samt mobilnätsstöd för fjärrstyrning; 4) videoöverföring för fjärrkörning; och 5) lagar och föreskrifter gällande fjärrstyrning. Innehållet i denna rapport avser att täcka tekniska detaljer och fynd inom samtliga av dessa ämnen, den är dock inte avsedd för att innehålla samtliga detaljer och resultat som redan har publicerats som vetenskapliga artiklar. Denna rapport bör ses som ett komplement till tidigare publicerade resultat.

A remotely operated road vehicle (RORV) refers to a vehicle operated wirelessly from a remote location. In this paper, we report results from an evaluation of two safety mechanisms: safe braking and disconnection. These safety mechanisms are included in the control software for RORV developed by Roboauto, an intelligent mobility solutions provider. The safety mechanisms monitor the communication system to detect packet transmission delays, lost messages, and outages caused by naturally occurring interference as well as denial-of-service (DoS) attacks. When the delay in the communication channel exceeds certain threshold values, the safety mechanisms are to initiate control actions to reduce the vehicle speed or stop the affected vehicle safely as soon as possible. To evaluate the effectiveness of the safety mechanisms, we exposed the vehicle control software to various communication failures using a software-in-the-loop (SIL) testing environment developed specifically for this study. Our results show that the safety mechanisms behaved correctly for a vast majority of the simulated communication failures. However, in a few cases, we noted that the safety mechanisms were triggered incorrectly, either too early or too late, according to the system specification.