Road freight transport is believed by many to be the first transport domain in which driverless (DL) vehicles will have a significant impact. However, in current literature almost no attention has been given to how the diffusion of DL trucks might occur and how it might affect the transport system. To make predictions on the market uptake and to model impacts of DL truck deployment, valid cost estimates of DL truck operations are crucial. In this paper, an analysis of costs and cost structures for DL truck operations, including indicative numerical cost estimates, is presented. The total cost of ownership for DL trucks compared with that for manually driven (MD) trucks has been analyzed for four different truck types (16-, 24-, 40-, and 64-ton trucks), for three scenarios reflecting pessimistic, intermediate, and optimistic assumptions on economic impacts of driving automation based on current literature. The results indicate that DL trucks may enable substantial cost savings compared with the MD truck baseline. In the base (intermediate) scenario, costs per 1,000 ton-kilometer decrease by 45%, 37%, 33%, and 29% for 16-, 24-, 40-, and 60-ton trucks, respectively. The findings confirm the established view in the literature that freight transport is a highly attractive area for DL vehicles because of the potential economic benefits.

Future sustainability impacts of driverless vehicles are subject to significant uncertainty which arise from complex systemic interactions within the transportation system and parallel social trends influencing transportation. One approach used to holistically address impacts of driverless vehicles is societal scenarios which capture and problematize the complex interactions. However, they are speculative in their nature and sensitive to the pre-conceptions and knowledge of the experts developing the scenarios. In this paper, multiple scenarios developed in several different studies are compared to create a deeper and broader understanding of system impacts of driverless vehicles and the future society with driverless vehicles than what is achieved through individual scenario studies. The findings show that there are four strategic uncertainties shaping the development: the role of the public and private sector, policy making for driverless vehicles, the impact of the sharing economy and the pace of driverless technology development. Most of the studied scenarios report higher traffic volumes than today. Impacts on social equity and the role of public transport vary significantly between the scenarios. Furthermore, the scenario studies expect the sharing economy to be an enabler to curb growth in travel volumes which is important if climate goals for transportation should be possible to meet. Further research efforts should address impacts of driverless vehicles in more systematic forms than societal scenarios but with wider system delimitations than in existing simulation studies.

The development of Self-Driving Vehicles (SDVs) is fast, and SDVs are predicted to have the potential to change mobility, human life, and society. Several positive and negative effects of SDVs are listed in the literature, but as the effects can be both counteracting and reinforcing depending on actions taken by different stakeholders, it is difficult to predict the outcome. In this chapter a scenario planning method is used to identify certain and uncertain trends, and to draw four different but plausible future scenarios for the development of SDVs. The four scenarios create a platform for policy discussions, development of regulations, and decision-making of different stakeholders. Most importantly, it shows that the development of SDVs depends on decisions made today.

In order to reliably predict and assess effects of congestion charges and other congestion mitigating measures, a transportation model including dynamic assignment and departure time choice is important. This paper presents a transport model that incorporates departure time choice for analysis of road users’ temporal adjustments and uses a mesoscopic traffic simulation model to capture the dynamic nature of congestion. Departure time choice modelling relies heavily on car users’ preferred times of travel and without knowledge of these no meaningful conclusions can be drawn from application of the model. This paper shows how preferred times of travel can be consistently derived from field observations and conditional probabilities of departure times using a reverse engineering approach. It is also shown how aggregation of origin–destination pairs with similar preferred departure time profiles can solve the problem of negative solutions resulting from the reverse engineering equation. The method is shown to work well for large-scale applications and results are given for the network of Stockholm.

This paper extends previous research by developing future scenarios for self-driving vehicles and their societal impacts in freight transport using Sweden as a case study. Freight experts from vehicle manufacturers, agencies, universities and hauliers were recruited for a workshop where they assessed the benefits, costs, possibilities and barriers for self-driving vehicles in freight transport. The paper shows that reduction in driver and vehicle costs, reduced number of incidents and more fuel-efficient driving are seen as the main benefits of self-driving vehicles in freight transport, and increased vehicle costs, lost jobs, reduced degree of filling and more transport as the main costs. Furthermore, reduced drivers’ costs, more hours-of-service and new business models are identified as the main drivers of the development and traffic management, small hauliers, loading and unloading and cross-border transport as the main barriers. The paper also integrates the description of possible developments of self-driving vehicles in freight transport into the four future scenarios developed for passenger transport in Sweden.

The intention with this report is to contribute toward the development of systemic and holistic studies of impacts of self-driving vehicles. The report is targeting system-level impacts of self-driving vehicles on the transportation system but also wider societal impacts on factors such as: land-use, public health, energy and emissions, etc. This report is complimentary to two papers that are focused on in-depth literature review of simulation studies  (Pernestål Brenden and Kristoffersson 2018) and future scenario studies of impacts of self-driving vehicles (Engholm, Kristoffersson, and Pernestål Brenden 2018).

The first aim of the report is to summarize knowledge to enable future design of a high-level conceptual framework for impacts from self-driving vehicles from a systems perspective. The second aim is to summarize knowledge on impacts from self-driving vehicles in a selection of the available literature. The main contributions of the report are the following:

• A terminology for different types of automated vehicles, connected vehicles and mobility concepts for automated vehicles is presented
• Frameworks for classifying system-level impacts from SDVs in the existing literature are summarized and analyzed
• Existing literature studies on system-level impacts from SDVs are synthesized and common themes and gaps in current research are analyzed

The terminology proposed in this report distinguishes between different types of automated and connected vehicles and is primarily intended as a tool to enable stringent analysis in this report when analyzing literature that apply different terminologies. Two frameworks for classifying system-level impacts are identified and compared. The analysis of the frameworks covers their scope, specification of mechanisms generating system impacts and briefly reviews their applicability as a starting point for developing a systems model of impacts from self-driving vehicles. The review of existing literature syntheses shows that there is a large variation in availability on literature for different system impacts. Impacts on road safety, road capacity and vehicle ownership forms are well studied. Examples of less studied impacts are costs of ownership, public health, infrastructure, air pollution and accessibility. The review identifies several contractionary mechanisms and effects that can affect various system-level impacts. The results of the review highlight the need to approach impact assessments of self-driving vehicles from a systemic and holistic point of view.

The charging systems in Sweden show that congestion charges can be an efficient (socio-economically beneficial) and effective policy measure for combating urban congestion. Furthermore, the technology of the Swedish charging systems has proven to work well, with high accuracy of correctly identified vehicles using the video technique with ANPR. The case of Gothenburg demonstrates this measure is not only less efficient if initial congestion levels are low, but also less efficient in the long run: the effects are declining in the long run. In Stockholm, the effects have increased over the years. The difference between the cities in this respect could be a result of the lower density city structure and high car dependence in Gothenburg. From this perspective, congestion charges are likely most successful in cities where congestion levels are high and where there exist good alternatives to driving. Chapter Contents: • 14.1 Introduction • 14.2 System designs • 14.3 Traffic effects • 14.3.1 Traffic volume across the cordon • 14.3.2 Traffic volume in the inner city • 14.3.3 Traffic volume on roads bypassing the inner city • 14.3.4 Travel times • 14.3.5 Long-term effects and effects of increased charging levels • 14.4 Adaptation strategies • 14.5 Revenues and system costs • 14.6 Model predictions • 14.7 Cost-benefit analysis, equity effects and company cars • 14.8 Public support • 14.9 Political support • 14.10 Lessons learnt and recommendations for other cities • References.

Under det senaste decenniet har nya datakällor, så som GPS‐data från taxibilar och storskaliga system av fasta detektorer, gett betydligt större möjligheter att kartlägga hur trängseln varierar i en stad, d.v.s. variation mellan gator och områden, olika tidpunkter på dagen och mellan olika månader eller år.

På den teoretiska sidan har det, under ungefär samma tidsperiod, upptäckts ett samband mellan fordonstäthet och hastighet på områdesnivå, vilket kallas det makroskopiska fundamentaldiagrammet (MFD). Tidigare har detta samband uppmätts på länknivå och kallas då fundamentaldiagram (FD). MFD kopplar samman antalet fordon i ett område med den genomsnittliga hastigheten ellerflödet i området. Man har också visat att MFD under ideala förhållanden är enegenskap hos nätverket i sig (infrastruktur och trafikstyrning), d.v.s. det beror inte på efterfrågan.

I denna rapport använder vi dessa nya trafikmätningsmetoder och teoretiska framsteg inom MFD för två syften. För det första beskriver vi hur trängseln varierar över dagen på Södermalm och i City‐området i Stockholm genom att titta på MFD från empiriska datakällor så som GPS‐data från taxi‐bilar, slangmätningar och restidskameror. För det andra jämför vi simulerat MFD för City‐området med empiriskt MFD för samma område. Detta för att validera hurväl City‐modellen framtagen med simuleringsverktyget Transmodeler kan återskapa trängselsituationen på områdesnivå.

Rapporten visar att väldefinierade MFD existerar både för Södermalm och Cityområdet.MFD visar att hastigheten sjunker och fordonstätheten ökar undermorgonens och eftermiddagens rusningstimmar, men trängselnivåerna når inte den punkt där flödet börjar avta trots att fordonstätheten ökar (hyperträngsel). Det är således trångt i innerstaden under rusningstimmarna, men kapaciteten i nätverket räcker ändå till. De två stora lederna Stadsgårdsleden och Sveavägen visar dock tecken på hyperträngsel om fundamentaldiagram skapas separat för dessa leder.

Vidare visar rapporten att MFD har stor potential som verktyg för att valideraen simuleringsmodell. I rapporten jämförs MFD från City‐området i Transmodeler med empirisk MFD för samma område. Simuleringsmodellen överskattar flöde och hastighet vid låg densitet. Vid hög densitet ändras dockbilden och simuleringsresultaten underskattar flöde och hastighet. Det verkarsom att kapaciteten i nätverket underskattas, vilket ger högre trängsel imodellen än i mätdata. MFD från Transmodeller visar lägre flöden underavvecklingen av rusningen än under uppbyggnaden, både under förmiddag och eftermiddag, vilket inte syns i de empiriska data. Detta tyder på att det finns stora kö‐problem i simuleringsmodellen, vilket man inte ser tecken på i empiriskt MFD.

Over the last decade, new data sources, such as GPS data from taxis and large-scale systems of fixed sensors, have created new opportunities to understand how congestion varies in a city, i.e. variation between streets and areas, different times of the day and between months or years.

On the theoretical side, a relationship between vehicle density and speed at arealevel has been discovered around the same point in time. This relationship is known as the macroscopic fundamental diagram (MFD). Previously, this relationship has been measured at link level and is then simply called the fundamental diagram (FD). MFD connects the number of vehicles in an area with the average speed or flow in that area. It has also been shown that underideal conditions MFD is a property of the network itself (infrastructure and traffic management), i.e. it does not depend on demand.

In this report, we use these new traffic measurement methods and theoretical advances in MFD for two purposes. First, we describe how congestion varies over the day at Södermalm and in the City area of Stockholm by looking at MFD from empirical data sources such as GPS data from taxis, tube measurements and travel time cameras. Secondly, we compare the simulated MFD for the Cityarea with the empirical MFD for the same area. This is done to validate how well the City model developed with the Transmodeler software simulation tool canreproduce the congestion situation at area level.

The report shows that well‐defined MFD exist for both Södermalm and the Cityarea. These MFD show that speed decreases and vehicle density increases during rush hour of the morning and afternoon, but that congestion levels donot reach the point where flow begins to decrease despite the increase invehicle density (hypercongestion). It is thus crowded in the inner city during peak hours, but capacity of the network is still enough. The two major arterials Stadsgårdsleden and Sveavägen show however signs of hypercongestion when fundamental diagrams are created separately for these arterials.

Furthermore, the report shows that MFD has great potential as a tool for validating simulation models. The report compares MFD from the City area of Transmodeler with empirical MFD for the same area. The simulation model overestimates flow and speed at low density. However, at high density, thepicture changes and simulation results underestimate flow and speed. It seems as if the capacity of the network is underestimated in the simulation model, resulting in higher congestion in the model than in measurement data. MFD from Transmodeler shows lower flows during the dissipation of the queues at rush hour than during the build‐up, both in the morning and afternoon, which isnot reflected in the empirical data. This indicates that there are major queuing problems in the simulation model, which cannot be seen in the empirical MFD.

Currently, there is a great need for new methods to collect travel data. Traditional methods have considerable drawbacks and, at the same time, the models used to analyse the transport system require more and more detailed and high-quality data. An alternative method that stands out as very promising is to capture raw data from devices that can use any positioning technology (e.g., GPS, WiFi positioning, GSM, etc.), followed by transforming the raw data into meaningful travel data. Since most smartphones are equipped with various sensors that can be used to determine the location of the smartphone, and since smartphones are integrated in the daily life of most people, they provide an unprecedented opportunity for large-scale travel data collection. This method has a great potential to solve the problems related to the estimation of distance/travel time, geographic coding of departure/destination locations and forgotten trips and it will also provide a more detailed and extensive data set. In a recently completed research project the feasibility of replacing or complementing the traditional travel diary, with a suite of tools that make use of smartphone collected travel data has been evaluated. The advantages and disadvantages of the traditional method and the proposed method were studied. For a fair comparison, both methods have been tested in the same city, at the same time, and with the same respondents. To achieve the objectives of the project, MEILI, a system that consists of a smartphone application for capturing the movement of users and a web application for allowing the users to annotate their movement, has been deployed. The recruitment of respondents is a critical phase for traditional travel diaries and, as expected, this was the case also for the smartphone based method. A lesson learnt was that it is important to simplify the registration process as much as possible. In total 2142 trips were collected and annotated by 171 users. 51 of the users annotated trips covering more than a week. The experiences from the field trial shows that a smartphone based travel diary collection is a very useful complement to traditional travel diary collection methods since it appeals to a different age group and collects more detailed travel data for a longer period. The main findings of the paper are that smartphone based data collection is feasible, that the algorithms to save battery work well and that trips of the same respondent vary considerably depending on day of the week.