INTERNATIONAL European Friedrich-List-Prize Internationales Verkehrswesen (72) 3 | 2020 44 Electrification of road freight transport Potentials and challenges of catenary guided systems for distribution system operators Electric road system, Distribution system operator, Catenary hybrid truck, Electrification scenario, Road freight transport electrification, Electric vehicle The intention of the European Union in mitigating emissions within the road freight transport sector is currently benefitting the general notion of electrification. Technological solutions like catenary hybrid trucks (CHT) are containing direct implications for the energy system. Therefore, the role of the distribution system operator (DSO), which primarily acts as a “fuel” supplier and integrator of systems within an existing system, is particularly noteworthy. To uncover its future systemic tasks it is necessary to determine the strategic potentials and challenges of catenary guided systems (CGS) for DSOs. Adrian Gunter T he intention of the European Union and its member states in mitigating emissions, which can be attributed to the road freight transport sector, is currently benefitting the general notion of electrification. Countries feel impelled to test electrified drive concepts in order to comply with climate targets set. The technolog- Innovative transport systems For the 15 th consecutive time the European Platform of Transport Sciences - EPTS - awards the “European Friedrich-List-Prize”. The prize, dedicated to young transport researchers, is named to honour the extraordinary contributions of Friedrich List, the visionary of transport in Europe of the 19th century, being a distinguished economist and respected transport scientist committed to the European idea. The European Friedrich-List-Prize is awarded for out-standing scientific papers in each of the categories Doctorate paper and Diploma paper, addressing topics in the transport field within a European context. In 2020 in total 12 scientific works have been nominated and evaluated. The award will be conferred during the 18th European Transport Congress in Rostock, Germany, on 13 October 2020. The results will be introduced on the website www.international-transportation.com and in “International Transportation - Collection 2020” (October issue). In the following you find a random selection of this year’s submissions summarized in drafts. Rostock Photo: Julia Boldt/ pixabay European Friedrich-List-Prize INTERNATIONAL Internationales Verkehrswesen (72) 3 | 2020 45 ical options for transport electrification, which are macroeconomically feasible and cope with new and existing regulations, e.g. (EU) 2019/ 1242, are yet to be finally determined. Due to their high energy demand, electric high-duty vehicles such as long-haul trucks may require another source of energy beyond batteries. This could be achieved with range extenders or roadside (catenary) solutions for continuous charging, as it is currently demonstrated in so-called Electric Road Systems (ERS) in countries like Sweden, Italy, and Germany. The roadside electrification of the road-freight transport sector via catenary hybrid trucks (CHT) - conductive power transfer through an overhead line-infrastructure (OL-I) extended with an internal combustion engine, battery storage, or fuel cell - contains direct implications for the energy system. In the interaction between the transport and energy sectors, the role of the distribution system operator (DSO) is particularly noteworthy. At the present stage, DSOs can already be perceived as a centric institution for the distribution of “fuels” for private battery electric vehicles (BEVs), as well as their integration into the existing distribution network structure. Concerning the application of a catenary guided system (CGS), however, the identification of load-based implications, the development and dimensioning of grid infrastructures, as well as the integration of mobile loads have yet not been fully covered. Furthermore, the structural parameters and regulatory conditions, which are crucial for the precise definition of the market design and the associated roles of actors, are not ultimately defined. Objective Until now, the evaluation of the CGS within studies is predominantly directed towards economical, ecological, and technical assessments of CHTs in comparison with its technological alternatives. Existing energy-economic evaluations of CGS are primarily based on key figures such as additional energy consumption, load profiles, and regional distribution of loads. However, there are currently no dedicated studies regarding the direct implications of a CGS for DSOs, which are assessing the associated strategic potentials and challenges. Accordingly, the leading question was investigated: Which factors in a catenary guided environment are relevant to establish scenarios in order to determine the strategic potentials and challenges for DSOs with horizon 2030? Methods The classification of a CGS within the sphere of DSOs was based on the findings of extensive literature research, compared and reassessed with the appraisals of governmental and non-governmental representatives of the German energy sector within the technical, regulatory, and commercial departments. Specifically, a three-stage procedure was pursued to determine relevant factors with a political and regulatory, economic, and technological scope. In Stage 1) “Data sources and data collection”, primary and secondary data were collected in the form of literature research, expert interviews, and discussion groups. Altogether, 17 interviews were conducted, which were deductively and inductively categorized and qualitatively und quantitively classified with the MAXQDA software. The gathered potential influencing factors were deductively attributed to socio-ecological, technical, economic, and political (STEP) perspectives and further reduced to relevant influencing factors via an internal factor assessment. In Stage 2) “Development of scenarios”, the further reduction of influencing factors into key factors was based on the application of an influence matrix, enabling their deductive categorization into superordinate categories. Subsequently, a delimitation and attribution of associated descriptors (characterization of key factors) and quantitative development paths (meta-analysis) were made. Based on the identified key factors, descrip- Figure 1: Categorization of the identified key factors and their descriptors Source: own visualization INTERNATIONAL European Friedrich-List-Prize Internationales Verkehrswesen (72) 3 | 2020 46 tors, and development paths, four heterogeneous scenarios were developed. In Stage 3) “Evaluation of the scenarios”, the different scenarios were compared and individually assessed based on a four-step procedure, comprising entrepreneurial evaluation and normative characterization methods. Results In the following, references are made to the findings that have been obtained in the course of empirical research and the analysis of the scenarios. Findings of the empirical research In absolute numbers, 32 potential influencing factors were identified, which have allowed a preliminary quantitative assessment. Accordingly, the political-regulatory aspects were highly relevant, accounting for 63 %. This can be explained by the fact that the technical (19 %) and economic (18 %) factors - regarding CGS - are subject to strong influences, combining merely minor implications (active/ passive ratio) on superordinated aspects. Through the application of a strategic early warning indicator system for DSOs, the number of factors could be further reduced to 20. In the further development process, 13 key factors were identified and assigned to the existing categories, allowing the development of four different scenarios and their key differentiators (parentheses), which are outlined in figure 1. Surrounding factors and external factors are reflecting the political and regulatory objectives, which are connected to the controlled implementation of energy transformation measures. The electrification of longhaul traffic, so far, has been influenced or even dominated by political and regulatory factors. With a high quantity, the German government’s climate targets were named as the decisive key factor. In this connection, the associated structural implications in the context of the energy industry and its potential of influence as well as the digitalization of the energy system in the context of network-related services were highlighted. Also, the acceptance for infrastructure projects was underlined, which should be communicated as a leverage effect to reduce emissions. Therefore, policy measures such as CO 2 -pricing or funding regimes (financing of OL-I) were identified as necessary instruments, mainly due to their regulating characteristics, as well as the influence on the ramp-up of electrified passenger and road-freight vehicles, which in return are crucial in raising the potentials of energy efficiency in the transport sector. Consuming factors were primarily identified based on developments in the areas of technology, volume, and structure of transport, as well as energy efficiency. Subsequently, the originating efficiency targets for the transport sector were regarded as essential to reduce the increased traffic volume and structure and thus the demand for conventional primary energy sources. Producing factors, which are based on the energy transition objectives and the regulatory frameworks, are considering the structures and capacities of renewable and conventional powerplants, as well as the development of storage technologies. Findings of the scenario analysis The adoption of the aforementioned factors and their translation into scenarios allows the identification of potentials and challenges for DSOs, which are explained in table 1. In the assessment of the scenarios, it was found that the potentials for regulated DSOs in the current market run-up (cf. Reference Scenario 2030) are severely limited by political and regulatory factors. This can be substantiated with the currently undetermined market design - regulated or market-based - the specification of operator models, the absence of dedicated billing models, and the associated financial structures of the OL-I, as well as the strong focus on battery-electric trucks. In this respect, the potential for DSOs can be identified in the acquisition of knowledge in the framework of closed projects, due to state-side financing and subsidies for the Photo: Siemens Mobility European Friedrich-List-Prize INTERNATIONAL Internationales Verkehrswesen (72) 3 | 2020 47 Scenarios Potentials Challenges Reference Scenario 2030 • Risk-free acquisition (financial) of competences in the field of technical management (maintenance, inspection) of OL-I • Practical investigation of load effects on the upstream network infrastructure and integration into network operation management • Continuing regulatory uncertainties regarding the type of network (currently customer installation), lack of technical standards • New technology field (OL-I) requires the acquisition of competence (workload of employees) Conservative Scenario 2030 • Transfer of gained knowledge (planning and consulting) from pilot projects to OL-I projects in other relevant European countries (e.g., Italy, Hungary or Poland) • Loss of systemic relevance to upstream grid operators (TSOs) due to stagnating RES expansion, loss of conventional power plant capacities and the procurement of services on the European capacity market Compliance Scenario 2030 • Extension of the technical management by commercial aspects (measurement and billing) as a service provider • Development of the service sector (establishment of OL-I) • The high commitment of human and financial resources with high financial risk (advance payment for the development of OL-I) • Planning, approval and acceptance risks; availability of operating resources (substations, etc.) Decarbonization Scenario 2030 • Acquisition of new concessions in the form of OL-I (compensation for expiring ones); expansion of the existing business area • Financing security through the application of the grid usage fee system (interest on capital employed) • New requirements for the integrated network planning of OL-I and the upstream distribution network • High investment and operating costs; submission to the regulatory regime (disclosure, cost review, etc.) Table 1: Potentials and challenges of the analyzed scenarios provision of services for the upstream infrastructure. Furthermore, the measurement of load and travel profiles could provide insights to DSOs, on how future grid expansions are to be dimensioned and planned accordingly. In principle, the existing capability profile can be used within this framework, which creates room for comprehensive integration of CGS into the network operation and management. In the context of a politically motivated market rampup of OL-I (cf. Decarbonization Scenario 2030), the participatory framework for DSOs would redefine itself. This is because the OL-I operator would then be determined in a bidding procedure based on tenders (marketbased) or in the form of a concession (regulated). Potentials would arise here primarily for the regulated area, which could operate outside its traditional network area with the construction and operation of OL-I. Nevertheless, regulatory authorities would have to decide on the legal form of the OL-I to determine the general framework and, therefore, its refinancing instruments. As an example, network fees could be mentioned here: It should be discussed whether higher returns can be generated from the participation in OL-I than in the existing network business so that the regulated DSO can operate in an economically sensible manner. Challenges can be identified in high leading times for system relevant infrastructure components, like transformer substations, limiting the possibility to react adequately to the ramp-up of OL-I. In this respect, integrated planning will be indispensable, which will potentially lead to a new and possibly greater complexity in the process of network planning. Accordingly, increased CAPEX and OPEX of OL-I and network infrastructure must be considered, which must be disclosed to regulators (publication of network charges) with the obligation to raise efficiencies. On this basis, there are risks in determining the right network charges to work cost-efficiently, but also to make the system interesting for consumers (CHTs). The lack of experience in measuring and billing of mobile loads, amplified by the division of the measuring mode (substation and CHT), are further increasing the complexity of reporting obligations to upstream network operators. Furthermore, it can be assumed that with the establishment of the infrastructure, a new focus will be put on the operators of infrastructures, as these are now daily visible in the form of OL-I. Conclusion In conclusion, it can be stated that the Decarbonization Scenario 2030 offers by far the highest potentials but also the highest challenges for DSOs. In the interest of the technological ramp-up, the economic compatibility, and the high level of complexity currently associated, it would be advisable to refrain from doing so at this stage. Within current pilot projects, DSOs might rather focus on the intelligence of the future OL-I and its control possibilities, which is strongly dependent on the regulatory framework, the degree of digitization, and social acceptance. For a realistic conclusion, the application of realistic operating phases is necessary to make the effects measurable outside the methodology. In the process of a further analysis, an extended cost analysis would have to be carried out, taking the individual threshold values and calculation rates into account. Nevertheless, the identified factors can form a common basis for European distribution system operators - as common European rules are in effect - thus allowing the framework of the scenarios to be adapted to individual conditions. ■ Adrian Gunter Student, Nuertingen-Geislingen University (HfWU), Nuertingen (DE) adrian_gunter@t-online.de INTERNATIONAL European Friedrich-List-Prize Internationales Verkehrswesen (72) 3 | 2020 48 Costs of operational hindrances Reduction of railway system costs by means of a holistic approach Railways, Infrastructure, Operation, System view In a globalized world mobility is a necessity for economic growth and a satisfied population. There are different possible transport modes, each with advantages and disadvantages compared to the others. The railway system stands for environmentally friendly transport for goods and passengers. To achieve future environmental requirements, it is necessary to maximise the performance of the railway system in a micro-economic term. A good way to ensure this is to reduce costs by finding a global cost optimum for the whole railway sector by assessing infrastructure and operational costs together. Markus Loidolt A consideration of all system costs of the railway system also includes the costs arising from unavailability. To assess the costs of unavailability it is necessary to include costs of all organizations involved. Looking at the organizational structure of a completely liberalised railway system, there are at least three large players [1]: The infrastructure manager (IM), railway undertakings for passengers (RU Passenger) and railway undertakings for freights (RU freight). Almost all additional costs of unavailability, such as additional operating costs, negative market reactions and penalties, are borne by the railway undertakings, the IM has hardly any additional costs to bear. Unfortunately, there are many scenarios in which the IM causes operational disruptions, but the subsequent costs must be borne by the RU. In this case, the costs of unavailability are external costs for the originator. Without any further regulations there are no incentives for the IM to reduce these costs, because doing so means higher costs for the own organization. Considering the total system costs, however, a cost optimization can only be achieved, if the IM includes external costs in its planning activities. As an illustrative example, the interactions between maintenance and permanent slow orders: One goal of an IM is to keep the costs of maintenance as low as possible. In a liberalized railway system and the resulting separate budget situation, the IM has the opportunity to save maintenance money by simply not carrying out maintenance tasks. The consequential costs, for example a disturbed railway operation due to permanent slow orders, must be borne by the RUs and are therefore outside the IMs budget. However, in the system’s view this leads to significantly higher costs, as the additional operational costs exceed the “savings” in maintenance. In extreme cases, the railway loses its justification for existence, weakened by permanent slow orders. Taking the costs of non-availability into account, ÖBB evaluated on strategic level that permanent slow orders are not a system compliant option, at least for the core network. [2] To guarantee the consideration of the costs of unavailability into the decision process of the IM, two conditions must be met: 1. Costs of unavailability must be assessed and monetised and 2. The costs must be included in the IM cash flow. Monetised costs of unavailability With the calculation scheme for costs of operational hindrances (CoH) [2], a tried and tested evaluation model for unavailability costs already exists at ÖBB. This is used for the development of strategies (speed restrictions, length of construction section, day work vs. night work, etc.) [3]. Table 1 is showing all negative cost effects that are included in the CoH system and the respective monetary valuation. Since the system has been in use since the 1990s (the underlying cost rates are updated annually), it can be considered stable. However, some boundary conditions have changed since the implementation, these should be integrated into the system as an update. For example, energy charging for trains in Austria has changed considerably since then. Soon, energy charging will be sharp for all trains, which means that the energy actually consumed is/ will be charged (in the 1990s, energy was charged at average cost rates). This kind of energy calculation means that a change in driving characteristics due to disturbances (e.g. slow orders) has a direct impact on the energy costs of the train ride. So, because of the new energy charging scheme, the energy consumption should be separately included into CoH for a more accurate cost representation. Nevertheless, the actual CoH- Positions considered Monetary evaluation delays and follow up delays • variable staff costs • variable costs of rolling stock - time based train rerouting • variable staff costs • variable costs of rolling stock additional trains • full costs of train operaton (without dep.) addtiional bus services • full costs of bus operaton (without dep.) additional shunting costs • variable staff costs additional operating costs • variable staff costs cancelling of trains • abolition of variable train costs • train parking costs negatibe market reaction • less customers, fee-repayments other costs • specifics Table 1: Negative cost effects included in the CoH system and the respective monetary valuation European Friedrich-List-Prize INTERNATIONAL Internationales Verkehrswesen (72) 3 | 2020 49 system is a cost-based model to assess unavailability at a reasonable cost level. The unavailability can therefore be assessed in monetary terms. However, for these costs to be considered in day-to-day planning activities, they must be included in the IMs cash flow. Otherwise they are dismissed as nonexpenditure costs, which is understandable from the point of view of the IM. However, from a system perspective, these costs must be taken into account, so that the costs of the railway system, and not just the costs of the IM, can be optimised. Performance Regime In order to make the railway system more attractive, the EU laid the legal basis for the performance regime (PR) with Article 11 in Directive 2001/ 14/ EC and Article 35 in Directive 2012/ 34/ EC. Quote from Article 35(1) of Directive 2012/ 34/ EC [4]: “Infrastructure charging schemes shall encourage railway undertakings and the infrastructure manager to minimise disruption and improve the performance of the railway network through a performance scheme. This scheme may include penalties for actions which disrupt the operation of the network, compensation for undertakings which suffer from disruption and bonuses that reward better-than-planned performance.” This EU directive must be transposed into national law by all member states of the European Union. The basic principle of the performance regime can be summarised as “organizations causing delays must pay penalties” and/ or “organizations that perform well receive discounts”. The infrastructure usage fee is used for billing. As in a liberalised railway system Track Access Charges (TACs) are the only link between IMs and RUs, the PR could be more than just an incentive for punctuality, but also a planning tool for the IM, leading to decisions that suit the system best. However, this only applies, if the cost rates of the performance regime are cost-based and not fixed for political reasons. As described above, CoH are exactly such cost-based cost rates. A comparison of the monetary evaluation of delay minutes of different PRs of the member states with ÖBB’s CoH, carried out in the course of the master thesis “Costs of Operational Hindrances”, leads to two conclusions: 1. There are big differences regarding the cost levels of the different PRs 2. The cost level of CoH is very well in line with the trend of PRs. While some IMs seem to have calibrated their PRs very well and have integrated the costs actually incurred by delays as a cost rate for evaluation (the extent to which this allocation has actually taken place is another matter), other schemes have obviously integrated cost rates according to political or any other interests. If the cost rates are too low, the incentive for punctuality is hardly achieved; as a planning tool for IM, PR is definitely not suitable in this case. The consequences of cost-based cost rates Since the majority of the monetary consequences of unavailability are borne by the RU, while the IM has hardly any additional costs, a common cost rate for both parties is not plausible. Rather, a cost-based definition must lead to two different cost rates, a very low one for the RU and a higher rate for the IM. If the IM causes a delay for example, it needs to compensate the additional expenses of the RU (high cost rates). If a RU causes a delay, it must bear most of the consequential costs anyway, in addition it must compensate the low additional costs of the IM (low cost rates). Additional arrangements must be made, if one RU is to blame for the delays of another RU. It is possible for the IM and RU to cancel the contract, as they are in a contractual relationship through the infrastructure usage charge. However, two independent RUs do not necessary share a common contractual basis. In order to counteract this problem, the idea of the so-called “star model” [5] was presented in one European state. This can serve as a model for other states. In the schematic model shown in figure 1, the IM, as the only party guaranteed to have a contractual relationship with all RUs, acts as an intermediary. The IM increases the costs for the RU1 causing the delay by the consequential costs of the additional operational expenditure of RU2, RU3 and RU4. This is calculated from the cost rates, which are generally charged, if the IM is at fault, as the RUs concerned also must bear these high consequential costs. The TACs of the RUs concerned will be reduced in total by the additional amount paid by RU1 as penalty payment. This means that the IM does not incur any costs nor is there an increase in budget. Summary In order to meet future requirements in both mobility and environmental terms, it is necessary to minimise the costs of the railway system and thus strengthen the position of the sector. Savings, such as investments in lower quality components or postponing maintenance measures, are only a short-sighted solution, which ultimately lead to higher costs. A far more sustainable method is to find a global cost optimum between infrastructure and operating costs by considering both cost positions when making decisions, which requires a monetary assessment of unavailability. The CoH calculation scheme acts as a cost-based model for this assessment. Using the calculation scheme developed in 2006 by the Institute for Railway and Transport Economics in cooperation with the Austrian Federal Railways, the costs resulting from Figure 1: The idea of the so-called “star model” [5] INTERNATIONAL European Friedrich-List-Prize Internationales Verkehrswesen (72) 3 | 2020 50 operational consequences can be calculated in dependence of the train type. For these costs to be considered in planning activities, they must also appear in the cash flow of the planning organization and not just disappear in a vacuum as so-called non-output-related costs. Even if these costs often occur in another organization, they form a relevant cost position and thus weaken the railway system. The performance regime, introduced by the EU legislation to reduce delays, is one way of capturing the costs of unavailability and can therefore be a planning tool for proper decisions. However, this planning tool only leads to system compliant decisions, if the rates meet the occurring costs in average. This condition can be fulfilled with CoH. They can serve as a good basis for a performance regime that provides incentives that benefit the railway system. Since an actual implementation of the presented concept leads to lower overall costs in the railway system, but increases the expenditure of the IM, an implementation can only go along with changes in the financing of the railway system. ■ REFERENCES [1] Marschnig, St. (2016): iTAC - innovative Track Access Charges, Graz University of Technology [2] ÖVG Band 55 - Spezial, Fahrweginstandhaltung auf Basis von Life-Cycle-Cost Berechnungen, 2002/ 08 [2] Veit, P.; Petri, K. (2008): Betriebserschwerniskosten - ein Baustein zur Systemoptimierung; ZEVrail Heft 5 [3] Marschnig, St.; Veit, P. (2010): Optimierte Einheitskosten - Sperrpausendauer und Baustellenlänge; ZEVrail Heft 10 [4] Directive 2012/ 34/ EC of the European Parliament and of the Council of the European Union; Official Journal of the European Union - 21. November 2012 [5] Office of Rail and Road, Performance Regime, April 2019 Markus Loidolt, Dipl.-Ing University Project Assistant, Institute of Railway Engineering and Transport Economy, Graz University of Technology, Graz (AT) markus.loidolt@tugraz.at Network traffic anomaly detection in IoT IoT, DDoS, Network anomaly, Machine learning, Logitboost, Cybersecurity The development of a public, packet-oriented communication network (Internet network), accompanied by an increase in the number of users and information and communication (IC) services, has also increased the amount of data transferred. Data stored, processed, and transmitted through the IC system is often the target of illegitimate users whose goal is to gain unauthorized access or to prevent legitimate users from accessing IC system resources. This results in an increase in the need for research in the field of IC protection in recent decades. Ivan Cvitić T he goal of protecting an IC system is to achieve and maintain the required level of basic security principles. The basic principles of security are presented by the CIA model, which embraces the integrity, confidentiality, and availability of IC resources. The availability principle is defined as the probability that the requested service (or other IC system resource) will be available to a legitimate user at the required time. There are several factors to impact the availability of IC resources negatively. They can be classified according to the source of these factors with the steadily increasing trend over the last ten years is network-oriented Distributed Denial of Service (DDoS) attack, or DDoS traffic as a means of conducting attacks and generating network traffic anomaly. Network traffic anomaly detection is a dynamic and broad area of research. Any network traffic pattern that deviates from the sample of a previously defined profile of legitimate (normal) traffic and has the potential to disrupt the normal operation of the IC is considered an anomaly. The legitimate traffic profile is defined by the values of traffic features recorded over a period of time in which the traffic generating terminal device is not security compromised and operates in the manner defined by the manufacturer. The root causes of network traffic anomalies may be related to performance or IC system security. One of the growing causes of securityrelated network traffic anomalies is DDoS attacks. This type of attack utilizes a number of compromised terminal devices to generate legitimate, DDoS traffic to the destination. The consequences of DDoS attacks are the degradation of quality or complete unavailability of IC services to legitimate users. The emergence of the Internet of Things (IoT) concept as a new direction of technological development and a new communication paradigm that brings together billions of new devices connected to the Internet, creates a new space of security vulnerabilities that can be exploited for unauthorized and malicious activities. The continuous growth in the number of such devices, their inadequate protection and the ability to generate traffic on the network, makes them ideal candidates for the creation of a botnet network to generate DDoS traffic of unprecedented traffic intensity. The concept of a smart home as one of the fastest-growing application areas of European Friedrich-List-Prize INTERNATIONAL Internationales Verkehrswesen (72) 3 | 2020 51 the IoT concept is becoming one of the most heterogeneous application areas in terms of the number of IoT devices manufacturers. Such devices are often delivered with minimal or no protection, and their security is also reduced by the ease of use required by end-users. Those users often do not have the adequate level of knowledge required to install and operate such devices. All of the above listed smart home devices are among the most vulnerable to many security threats, emphasizing the use of such devices to generate DDoS traffic. The concept of IoT offers numerous benefits in different fields of application, but from the point of security view, it also highlights many challenges that need to be adequately addressed. Research within this doctoral thesis considers the smart home environment as one of the fastest-growing application areas within the IoT concept. Devices within this environment have many limitations and disadvantages that make them potential generators of DDoS traffic. According to predictions, by the end of 2020, approximately 31 billion IoT devices will exist globally, and 75 billion till 2025. In this case, 41 %, or 12.86 billion IoT devices will be installed within the concept of a smart home (SH). The limitations of IoT devices in general, and thus SHIoT (smart home IoT) devices, are described in the previous researches, covering hardware constraints, high autonomy requirements, and low cost of production, which reduces the ability to implement advanced security methods and increases the risk of numerous threats. Traffic generated by SHIoT devices or MTC (Machine Type Communication) traffic is different from traffic generated through conventional devices, HTC (Human Type Communication) traffic. Although SHIoT devices are characterized by heterogeneity, MTC traffic is homogeneous in contrast to HTC traffic, which means that devices of the same or similar purpose behave approximately equally, that is, generate traffic of similar characteristics. The identified shortcomings of previous research, such as taking into account of SHIoT traffic features when detecting DDoS traffic, the consideration of classes of SHIoT devices that generate roughly equal values of traffic features, and the number of devices used in the study, will be sought to be remedied by planned research. The importance of this research is also evident through the increasing number of research and projects in this field. An example of this is the project called Mitigating IoT-Based Distributed Denial Of Service (DDoS), implemented by NIST (National Institute of Standards and Technology) and NCCoE (National Cybersecurity Center of Excellence), which addresses the issue of generating DDoS traffic through an IoT device. Trough research within this doctoral thesis, the laboratory environment of the smart home was formed and shown in figure 1. Such an environment is comprised of a variety of SHIoT devices, along with an accompanying communications infrastructure and softwarehardware platform that enables traffic collection and data set to be applied in later stages of research and development of network traffic anomaly detection models. In addition to the primary data collected through the process described above, the research also included secondary data, encompassing a greater variety of SHIoT devices. The reason for this is the heterogeneity of devices that can exist in the observed environment. A total of 41 devices in a smart home environment were used for this doctoral research. According to statistics, there are differences in the estimation of the average Figure 1: Smart home laboratory environment INTERNATIONAL European Friedrich-List-Prize Internationales Verkehrswesen (72) 3 | 2020 52 number of SHIoT devices per household that has a certain form of smart home implemented. These estimates range from 6.53 to 14 SHoT devices per household. In the Republic of Croatia, smart home representation is still low, and telecom operators are assuming the role of smart home providers through the offering of end-user SHIoT devices. For example, Iskon Internet service provider offers customers the option of purchasing a smart home package that makes four SHIoT devices, while telecom operator A1 provides users with the ability to deploy a total of five SHIoT devices in a smart home environment. Despite mentioned, this research sought to achieve the greatest possible variety of SHIoT devices due to the need to define device classes based on the characteristics of the traffic generated. Therefore, the number of devices used is higher than the current statistical estimate of the average value of SHIoT devices per smart home in the Republic of Croatia and worldwide. The work of the developed model of detection of illegitimate DDoS traffic takes place in two stages. The first phase is a prerequisite for the later detection of DDoS traffic in the second phase of operation and implies the classification of the SHIoT device based on the generated traffic flow. One of the basic metrics that indicate model performance is classification accuracy and kappa statistics. According to the classification accuracy, all models show high performance, which means that based on the observed flow, they can determine with high accuracy whether the traffic flow is the result of legitimate device communication or the device generates DDoS traffic. Research has shown that it is possible to define device classes based on the variation of the received and sent traffic ratio, and it is possible to classify devices into defined classes based on the traffic flow features such devices generate. Finally, depending on the affiliation of an individual device to a defined class, it is possible to determine whether the traffic flow that the device generates is an anomaly in the form of DDoS traffic or legitimate traffic. The research carried within this doctoral dissertation is of significant importance for the development of the research area since it considers the challenges of the fastgrowing and omnipresent IoT concept. This concept represents a new paradigm of the application of information and communication technologies and services where the devices in such an environment may become generators of network traffic anomalies and cyber-attacks. Previous research very rarely took into consideration such an environment during the research of network traffic anomalies detection in spite of the fact that the cyber-attack generators, such as DDoS, in the last five years have been the devices under the IoT concept. Accordingly, the focus of network traffic anomalies detection research in the IoT concept represents an important step also for future research of this area and highlighting the importance of anomalies detection in such an environment, which is showing a strong growing tendency. The possibility of applying the results of this research in practice is seen from the aspect of several stakeholders such as the end-user, Internet service provider (telecom operator) and device manufacturer, and service provider in smart IoT environments. From the aspect of the user as a smart home stakeholder, the need is emphasized for the devices to function in the way planned by the manufacturer, i.e., that all the device functionalities are available in the requested time. The generation of DDoS traffic can also cause unplanned behavior of the device that generates such traffic, which may reduce its functionalities or make the device fully inaccessible. Therefore, it is in the interest of the user to make timely detection of the unwanted behavior of the device, which enables activities that follow the detection. The generation of DDoS traffic by means of a large number of IoT devices in the smart home environment may negatively impact also the network and the server infrastructure of the telecom operator. Since telecom operators are often also the smart home service providers, it is in their interest to timely detect unauthorized behavior of the device in order to protect their own network infrastructure. The manufacturers of such devices have to ensure proper operation of the devices in order to increase the satisfaction of the users and the market expansion. They will ensure this by timely detection of the unauthorized operation of the device that will make it possible for them to respond to the unwanted events and to ensure the desired level of user satisfaction. ■ Ivan Cvitić, Ph.D. Assistant, Faculty of Transport and Traffic Sciences, Department of Information and Communication Traffic, University of Zagreb (HR) ivan.cvitic@fpz.unizg.hr Photo: Social Cut/ Unsplash Transforming Transport - at a glance Taken together in a spanning compilation: International Transportation - Collection 2020 presents the complete range of this year’s English contributions with the topic “Transforming Transport” plus additional articles in the sections • Strategies • Best Practice • Products & Solutions • Science & Research “International Transportation - Collection 2020” comes as an e-journal only - ready for desktop computers as well as mobile devices. Publishing date is 12 October 2020. For further information see the website: www.international-transportation.com European Friedrich-List-Prize INTERNATIONAL Internationales Verkehrswesen (72) 3 | 2020 53 Innovative transport systems and mobility services Integrating autonomous vehicles into the public transport system Altering transport system, Autonomous vehicle, Integration, Mobility service, System engineering The developments of infocommunication and vehicle technology have altered the passenger transport system and given way to the emergence of innovative mobility services. During a Ph.D. research, the author focused on this alteration. The altering transport system, the planning and operational processes of new mobility services, the impacts of mobility services based on autonomous vehicles, as well as the automation opportunities of planning and operational functions were examined from the viewpoint of transportation engineering. Dávid Földes T echnical innovations, such as automation, have facilitated sustainable mobility developments (e.g. transitional mobility services, such as car-sharing, which blur the borderlines between private and public transport). The objective of such developments is the efficient management of resources as well as complying with user preferences. Automation can enhance operational efficiency and traveller’s comfort. An automated system operates on clearly defined algorithms; an autonomous system is able to make decisions using its cognitive and self-learning abilities. As a result of technological developments, a smart mobility system can be introduced, which combines human knowledge, intelligence, and decision-making processes. Data and information have become key to decision-making. Consequently, the transport system can be considered as a special information system. A systematic revealing of elements and connections is required. Studies in automation focus on the control and traffic issues of Autonomous Vehicles (AVs) [1]. However, passenger handling, operation, and maintenance can also be automatized [2]. Placing AVs into a wider-approach within the passenger transport system has moderately been emphasized so far. Albeit transport modes are altering, new methods are required for planning, organizing and operating transport. A new type of mobility service based on small capacity AVs emerges, which is shared, on-demand and accessible only with advance ordering via a mobile application [3, 4]. Mobility becomes more and more a pre-planned activity requiring proactiveness from the travellers. Human skills, the traveller’s decision-making processes, and behaviour are also altering. Accordingly, the development of innovative information management methods and services supporting decisionmaking is required. Therefore, the objectives of the research were to model AV-based transport systems on an urban scale, as well as mobility and information services, moreover, to elaborate system planning principles and evaluation methods. The focus was placed both on the operation and the traveller. Since the object of transport is the traveller, revealing expectations towards new mobility services is especially important. If the travellers’ expectations are met, the adoption of new technology can be enhanced. This summary briefly summarizes the most relevant results of the research, namely the model of smart mobility, the alteration in mobility services, the information system model for the planning and operation of AVbased services and the complex automation levels. Methods The methods applied during the research are as follows. A special method for analysing and modelling information systems was developed and implemented, which reveals structural and operational relationships in different resolutions (break-ups). Furthermore, relational data modelling was used for the elaboration of the database structure for the operation of AV-based mobility service. Multicriteria analysis, which is appropriate for complex systems, was used to model the smart mobility system. Weighted Sum Model was applied to determine automation levels. In order to obtain the right conclusions about the expectations towards AV-based mobility service, preferences were collected by a questionnaire survey. The connections between data groups were examined to determine the impact of each data group on each other. Both deductive and inductive logic was applied to draw conclusions. Results Smart mobility New transport-related developments should be integrated into a system. This is called smart mobility, which is a decisive sub-system of the smart city; it realizes physical relationships between other sub-systems. It includes human knowledge, intelligence and a mechanism of decision-making applying information and communication technologies cooperating in transport infrastructure, in vehicles, and by travellers. The smart traveller is one of the smart mobility sub-systems and covers pedestrians, bikers, passengers and drivers as well. The structural and operational model of the smart mobility system focusing on the information management of the traveller was defined. The author found that the information management of a machine and a human are sim- INTERNATIONAL European Friedrich-List-Prize Internationales Verkehrswesen (72) 3 | 2020 54 ilar. The machine system can be developed according to the revealed attributes of human information management. Consequently, information management can be supported and even replaced by an adapted info-communication technology. Alteration in mobility services Based on the literature review and situation analysis, the author identified the alteration in transport modes (see in figure 1). The envisioned future modes were depicted in terms of the number of passengers per vehicle and flexibility. Flexibility is a complex indicator depending on several aspects (e.g. spatial accessibility). Transitional transport modes and, even more, the majority of private car use can be replaced by a shared, on-demand mobility service based on small capacity AVs accessible only with advance ordering via a mobile application. The types and the characteristics of this service were defined. Among others, a rather flexible door-to-door type and a slightly less flexible feeder type linked to a high capacity line were also distinguished. The feeder type may run on a fix route or according to a fix timetable. As the capacity of the built infrastructure is limited, travel demands can be served efficiently by shared and feeder mobility services. As AV-based mobility services are in an initial phase, the research elaborated on the structural and operational model of shared AV. The conclusion of the subresearch is that autonomy is a relative concept, since the coordination of several centres with different functions is required to plan, control and operate AV-based mobility services. That is why the integrated mobility management centre organizational unit was introduced with its defined tasks (e.g. management of operational data in an integrated database). There are several expected impacts of shared AV; the impact fields were identified, and a model was developed to calculate the alteration in modal share. Stated preferences are used as input data by the model. It was found that private car use could be significantly reduced by the introduction of a flexible, shared, AV-based mobility service. Planning and operation of AV-based mobility services The planning and operation of shared AV require new methods. The aspects that cause alteration in conventional methods are as follows: • more complex system structure, • new and unknown technology, • dynamism of the data and • travellers’ expectations towards more adaptive and sustainable service. Travelers should also perform existing tasks in a novel way or should solve new tasks as well (e.g. ordering, boarding, payment). The role of personnel is reduced, and the driver’s requirement can be ignored. New solutions are to be applied both in operation (e.g. charging) and in passenger handling (e.g. information provision). Functions with major alterations are realtime demand-capacity assignment, vehicle run planning, customized information services and vehicle charging. The information system model was defined for the planning and operation of shared AV. Considering travellers’ expectations is particularly important as the developments of such services are at an early stage. Accordingly, to define the model, the author determined the input data groups resulted from preferences and elaborated on the data collection method (questionnaire survey). It was found that travellers’ socio-demographic and mobility habits influence expectations towards the mobility service based on AVs. Automation levels The calculation method of complex automation levels was determined for road-based mobility services. Control functions, service planning and management, as well as passenger-handling functions were considered. Four levels of automation were distinguished. Applying the method, the automation level of a mobility service can be described in a general and simplified way using only one value (table 1). Figure 1: Alteration in transport modes o. name description the entity which makes decisions and executes 1 automation All processes are executed by humans. The human has full responsibility, there is no direct machine support. human 2 machine assistance Decision-making is supported by the machine. However, the role of a human is significant. human aided by machine 3 partial automation A significant part of the processes is executed by the machine. The personnel monitor the processes. mostly machine with human confirmation 4 full automation Processes are completely operated by the machine. The personnel attend only as a supervisor machine Table 1: Complex automation levels European Friedrich-List-Prize INTERNATIONAL Internationales Verkehrswesen (72) 3 | 2020 55 Automation impacts on the required human abilities. To determine the aggregated ability alteration, an assessment method was developed. The method considers every sub-function for the entire ride. It was found that the required human cognitive capability, all in all, decreases significantly as the consequence of automation and machine support, whereas requirements towards human abilities related to smartphone use rise. Conclusion The main contributions of this research were the developed models of smart mobility. Moreover, it was revealed and analysed the characteristics of smart traveller’s information management and shared AV mobility service. Furthermore, it was elaborated on the information system model for planning and operating shared AV. In addition, complex automation levels for road-based mobility services were determined, and the alteration in required human abilities analysed. The results can contribute to facilitating and preparing the alteration of the transport system and the integration of AV-based services. They were already included in the curricula of subjects at Budapest University of Technology and Economics.The most relevant key findings are as follows: • Information management can be supported and even replaced by info-communication technology. • Autonomy is a relative concept; coordination of several centres with different functions are required. • According to travellers’ preferences, private car use could be significantly reduced by the introduction of shared AV. • Less human thinking is required because of machine support. Moreover, the human can be replaced in certain functions by the machine. As automation technology is relatively new, experience is available neither from operators nor from travellers. Objective is to continue the development of evaluation methods for mobility services. The evaluation covers service quality, flexibility, features of integrity and automation, as well as customization. The research will continue in order to develop information services for supporting travellers’ decision-making and also to develop AV-based mobility services. ■ Special thanks to Csaba Csiszár, Ph.D. for providing guidance and feedback throughout the research as the supervisor. REFERENCES [1] Szalay, Zs.; Nyerges, Á.; Hamar, Z., Hesz, M. (2017): Technical Specification Methodology for an Automotive Proving Ground Dedicated to Connected and Automated Vehicles. Periodica Polytechnica Transportation Engineering, vol. 45, no. 3, pp. 168-174. [2] Chen, T.D.; Kockelman, K.M.; Hanna, J.P. (2016): Operations of a shared, autonomous, electric vehicle fleet: Implications of vehicle & charging infrastructure decisions. Transportation Research Part A: Policy and Practice, vol. 94, pp. 243-254 [3] Bansal, P.; Kockelman, K. M.; Singh, A. (2016): Assessing public opinions of and interest in new vehicle technologies: An Austin perspective. Transportation Research Part C: Emerging Technologies, vol. 67, pp. 1-14 [4] Winter, K.; Cats, O.; Correia, G.; van Arem, B. (2016): Designing an Automated Demand- Responsive Transport System: Fleet Size and Performance Analysis for the Case of a Campus-Train Station Service. TRB 95th Annual Meeting Compendium of Papers Dávid Földes, Ph.D. Research associate, Department of Transport Technology and Economics, Faculty of Transportation Engineering and Vehicle Engineering, Budapest University of Technology and Economics foldes.david@mail.bme.hu Creation of mobility packages based on the MaaS-concept Mobility as a Service, Carsharing, Transport mode, Mobility solution Urban mobility issues have impact on citizens’ quality of life and the overall sustainability of cities. Mobility as a Service (MaaS) is a new paradigm, which enables the increase of efficiency in passenger transportation networks. MaaS will integrate transport modes and mobility solutions with the emergence of new technologies. Kerényi Tamás I n today’s world there is a shift from rural areas to cities, where urbanization will have a significant impact on the way we travel around. Mobility issues will have impact on citizens’ quality of life and the overall sustainability of cities. It is necessary to improve the conditions of sustainable travel through development of vehicle technology, infrastructure and intelligent transportation systems, but also changes in travel behaviour are necessary in order to reduce private car dependency and the share of trips made with individual vehicles [1]. By introducing various solutions practitioners tried to change the habits of travelers to choose alternative modes of transport instead of their private cars. An intervention option is to motivate travellers through rewards and punishments, which are translated in the field of transportation to congestion charges, taxes and parking fees [2]. Encouragement is INTERNATIONAL European Friedrich-List-Prize Internationales Verkehrswesen (72) 3 | 2020 56 another type of intervention, where providing free public transportation passes is a suitable example [3]. In recent years another trend, the sharing economy and collaborative consumption is expanding in many industries. In transportation the number of car-sharing users is growing rapidly. While about 1,000 cities worldwide have a bike-sharing program today, ride-sharing services has expanded similarly. Using these new technologies the need arises to introduce new types of services in transportation [4]. Furthermore the digitalization has opened up new opportunities, which has enabled the development of some new mobility services, like multimodal travel information services, where travelers can choose between a high variety of different services and options. These services are usually limited to providing route planning and real time information [5]. Mobility as a Service (MaaS) is a new paradigm, which enables the increase of efficiency in passenger transportation networks. It is a solution that combines a number of services and provides a platform, where the intermodal journey planning and payment are integrated [6]. However it is an issue to predict user demand for these services, which requires modeling to support long term strategic decisions [7]. Furthermore it should be considered that the changes in mobility have shifted the role of public authorities, as they are providing the mobility services. Exploring of new operational models of mobility services is a challenge [8]. It is clear that the MaaS concept has to solve the problem of providing suitable mobility packages for the users, but MaaS research and development is still at an early stage and far from a definitely mature concept. More importantly there is a lack of knowledge, how the plans should be created and what local aspects could be taken into account. When creating a package, it is essential to specify what kind of transportation modes are included, to what extent the specific modes should be offered and what circumstances may have an effect on the package structure. In general the users can be offered by three types of monthly packages: fixed number of usage, flexible travels and unlimited option. The fixed package is predefined by the service provider and the traveler may choose only specific days of travel and pay by single usage, e.g. using car-sharing service on a chosen day. In case of the flexible option the traveler can customize the usage service to a limited extent, e.g. using 10 days of bike-sharing service within 30 days. The unlimited option covers a typical package for a predefined time of usage, when the service is available without any limitations, e.g. public transport monthly pass. Background MaaS is a new approach that will integrate transport modes and mobility solutions with the emergence of new technologies. The concept includes several until now separately handled services, such as planning, booking, payment and ticketing. Already in 2011 in an UITP position paper the association has forecasted that combining various transport modes, as car-sharing, taxi, shared taxis, bicycle and bike-sharing, car-pooling, demandresponsive transport can complement classic public transport [9]. The first definition appeared in 2014 by Hietanen: “A mobility distribution model in which customer’s major transportation needs are met over one interface and are offered by a service provider” [10]. This paradigm shift softens boundaries between different transportation modes with new services constantly being added. It offers travelers easy, flexible, reliable and sustainable choices solving urban mobility issues. Several implementations of different levels of MaaS services are present, which were designed by public and private operators [11]. One of the first MaaS concepts was realized by MaaS Global in Finland. The packages contain mobility services, offers personalized bundles for journeys and provide value added services. However, realizing such a service is complicated, and it can only be successful, if user needs are mapped and suitable mobility packages are provided, which reflect the real usage requirements. Method The purpose of the paper was to produce mobility packages. These packages support the efficient realization and support user demand based usage of the MaaS concept. Combining different services into one package is based on the idea that users value more grouped packages than individual items [12]. In order to perform the combination a method was elaborated, which takes into account city specific aspects and special features of transportation modes (figure 1). We have defined the most relevant modes of transport, which may be included in the mobility packages: public transport, bike-sharing, car-sharing and taxi. To each mode a level of package is selected based on different features. The features are combined from specific aspects, separately for each transportation mode. In order to analyze city characteristics those aspects were chosen, which represent relevant features in a city when considering mobility choices and are publicly available from different sources. The aspects are considered as quantifiable characteristics of cities. First the Figure 1: Steps of mobility package creation process European Friedrich-List-Prize INTERNATIONAL Internationales Verkehrswesen (72) 3 | 2020 57 general climate of cities is characterized by the sunshine hours and rainfall. Then financial information was collected, the average monthly salary, cost of living, public transport pass price and taxi tariffs. Specific indexes were also considered to receive information about the situation of transportation in the cities (traffic index, travel time index, PT satisfaction). The modal split of transportation modes were received from two different sources. The size of the agglomeration was compared to the size of the city, while the density of the city was captured, so that the number of commuters and citizens can be considered. Environmental awareness of the citizens was measured based on the amount of waste and commitment to climate change. Finally the median age and male/ female ratio was collected, so that general sociodemographic features can be defined. Mobility packages In order to create the mobility packages, the package levels were assigned to the predefined definitions and the final suggested packages were established regarding the different transportation modes (table 1). Thus suitable mobility packages were created for each selected city considering local aspects and user needs. Of course these packages are general ones, and cities should offer some bigger and smaller packages of specific transportation modes in order to fulfill all user requirements. As this method considers aggregated data, it may be used as a starting point when designing mobility packages. In order to design better packages, further and more detailed data acquisition, household surveys, stakeholder interviews and other information sources would be necessary. As a final result for example in Brussels a limited public transport package is suggested with 10 days of usage within a month, with one hour of bike-sharing within a day, and three hours and 20 km of taxi. In Budapest unlimited public transport would be suitable with three hours of bike-sharing, no car-sharing and pay-as-you-go taxi. Vienna is another typical example with unlimited public transport usage, unlimited bike-sharing, three hours of car-sharing and 10 km taxi. Conclusion The aim of this paper was to create mobility packages based on local characteristics of cities. 17 aspects were chosen, which represent cities. When creating the packages the following transport modes were considered: public transport, bike-sharing, car-sharing and taxi. The aspects were grouped to features, when considering different modes of transportation. As a result the characteristics of 15 European cities were collected and features were assigned, thus the combination of packages was suggested to each city. MaaS Global is the world’s first mobility as a service (MaaS) operator leading the change in how the world moves in the future. Whim app makes smart travelling easy by incorporating all transportation modes available - from public transport to taxi rides, car rental to city bikes and more into one service. Among the researched cities Helsinki is the only one where the traveler can choose from created mobility packages, so the results can be compared. For regular travelers Whim provides the following: unlimited local public transport, 30 min city bike, car rental just EUR 49 and taxi rides within a 5-km radius. The contents of the two packages are very similar, public transport is the same, bike-sharing and taxi is same but, in our package provides more 1 hour and 10 km. Whim has car rental service and in the research has carsharing service which cannot be properly compared. ■ REFERENCES [1] Graham-Rowe, E.; Skippon, S.; Gardner, B.; Abraham, C. (2011): Can we reduce car use and, if so, how? A review of available evidence, Transportation Research Part A, Vol. 45, pp. 401-418 [2] Karlsson, M. I. C.; Sochor, J.; Strömberg, H. (2016): Developing the ‘Service’ in Mobility as a Service: experiences from a field trial of an innovative travel brokerage, Transportation Research Procedia, Vol. 14, pp. 3265-3273 Taxi Car-sharing Bike-sharing Public Transport Brussels 20 km free per month unlimited pay-as-you-go 20 days per month Prague 20 km free per month one hour free per day pay-as-you-go 20 days per month Hamburg free 20 km free per month one hour free per day one hour free per day 20 days per month Athens 10 km free per month three hours free per day pay-as-you-go 20 days per month Budapest 10 km free per month one hour free per day pay-as-you-go 20 days per month Bucharest 10 km free per month one hour free per day pay-as-you-go 20 days per month Warsaw 20 km free per month one hour free per day pay-as-you-go 20 days per month Oslo 10 km free per month pay-as-you-go pay-as-you-go 20 days per month Sofia 10 km free per month one hour free per day pay-as-you-go 20 days per month Stockholm pay-as-you-go pay-as-you-go one hour free per day 20 days per month Glasgow 20 km free per month three hours free per day pay-as-you-go 20 days per month Wien 20 km free per month one hour free per day pay-as-you-go 20 days per month Copenhagen 10 km free per month pay-as-you-go unlimited 20 days per month Helsinki 10 km free per month one hour free per day one hour free per day 20 days per month Turin 50 km free per month unlimited pay-as-you-go 20 days per month Table 1: Complex automation levels INTERNATIONAL European Friedrich-List-Prize Internationales Verkehrswesen (72) 3 | 2020 58 [3] Cats, O.; Susilo, Y. O.; Reimal, T. (2017): The prospects of fare-free public transport: evidence from Tallinn, Transportation, Vol. 44, Issue 5., pp. 1083-1104 [4] Jittrapirom, P.; Caiati, V.; Feneri, A-M. Ebrahimigharehbaghi, S.; Alonso-González, M. J.; Narayan, J. (2017): Mobility as a Service: A Critical Review of Definitions, Assessments of Schemes, and Key Challenges, Urban Planning, Vol. 2, Issue 2, pp. 13-25 [5] Esztergár-Kiss, D.; Csiszár, Cs. (2015): Evaluation of multimodal journey planners and definition of service levels, International Journal of Intelligent Transportation Systems Research, Springer, September 2015, Vol 13, Issue 3, pp 154-165. DOI 10.1007/ s13177-014- 0093-0 [6] Kamargianni, M.; Li ,W.; Matyas, M.; Schäfer, A. (2016): A Critical Review of New Mobility Services for Urban Transport, Transportation Research Procedia, Vol. 14, pp. 3294-3303 [7] Meurs, H.; Timmermans, H. (2017): Mobility as a Service as a Multi-Sided Market: Challenges for Modeling, 96th Transportation Research Board (TRB) Annual Meeting, Washington, United States, 8-12 January 2017 [8] Li Y., Voege T. (2017): Mobility as a Service (MaaS): Challenges of Implementation and Policy Required, Journal of Transportation Technologies, Vol. 7, Issue 2., pp. 95-106 [9] UITP (2011): Becoming a Real Mobility Provider Combined Mobility, position paper, www.uitp.org/ sites/ default/ files/ cck-focus-papers-files/ FPComMob-en.pdf [10] Hietanen, S. (2014): “Mobility as a Service” - The new transport model? , Eurotransport, Vol. 12, Issue 2, pp. 2-4 [11] Sochor, J.; Arby, H.; Karlsson, M. (2017): The topology of Mobility as a Service: A tool for understanding effects on business and society, user behavior, and technological requirements, 24th World Congress on Intelligent Transportation Systems, Montreal, Canada, 2017. October 29 - November 2 [12] Enoch, M. (2012): Sustainable Transport, Mobility Management and Travel Plans, Routledge, Taylor & Francis Group Ltd Kerényi Tamás Transport Development Junior Project Manager, BFK Budapest Development Center Nonprofit Ltd., Budapest (HU) kerenyi.tamas@icloud.com Non-probability recruitment strategies for innovative smartphone-based travel surveys Survey, Travel behaviour, Non-probability sampling, GPS, App Pros and cons of non-probability sampling are varied: Cost-effective techniques enable targeting specific population groups, flexibly reacting to changes in sample structures, and increasing participant motivation. Yet representativeness is frequently doubted when the principles of probability samplings are violated. In the City of Dresden, a travel survey was conducted using a tracking app called TravelVu. This article assesses the performance of both broad-based and individually tailored recruitment strategies, accessing different resources (e.g., news, social media, local ads, printed materials). Johannes Weber, Stefan Hubrich, Rico Wittwer, Regine Gerike A n important basis for urban and transport planning are (household) travel surveys. They provide information on how a transport system with all its interacting modes is currently used and, when repeatedly conducted, which trends are emerging over time. One of the main challenges facing large-scale travel surveys is the fact that response rates are declining — this can be observed in Germany and beyond. Consequently, risks of selectivity as well as costs and efforts for obtaining high-quality data increase [1]. In addition, innovative tools such as smartphone apps have been coming to the foreground which create new possibilities for data collection via GPS-tracking. Unlike previous data-collecting methods, tracking apps collect data in real time, lowering the overall respondent burden and offering quality framework for a longitudinal survey design [2, 3]. In light of growing interest in new data-collection methods and their integration into traditional survey designs for enhancing data content, a further question arises: Do more targeted sampling and recruitment strategies exist for obtaining adequate sample sizes within an acceptable quality and cost range? Non-probability samples offer various promising approaches such as in-street recruitment, distribution within workplaces, and the use of social media; yet they also entail challenges regarding systematic sample losses, representativeness, and sample bias [4]. Compared to other fields of research, there is minimal experience in applying such non-probability sample methods to traditional travel surveys [4]. Following up on this, the thesis - submitted to the European Friedrich-List-Prize 2020 - conducted a major travel survey in the City of Dresden (managed and supervised by the TU Dresden), with a digital travel survey app called TravelVu and a non-probability recruitment concept. The goal was to learn about the performance and effects of different recruitment approaches in terms of sampling composition, costs and survey response. Embedded in a research co-project named “Travelviewer - data for lowcarbon sustainable transport systems” financed by EIT Climate-KIC, specific travel surveys were conducted in three other European sites, demonstrating the use of TravelVu and testing new means of recruitment. Non-probability sampling in market and public opinion research Particularly in market and public opinion research, nonprobability sampling has the critical advantage of producing cost and expense savings in comparison to regis- European Friedrich-List-Prize INTERNATIONAL Internationales Verkehrswesen (72) 3 | 2020 59 ter-based sampling. However, certain sample biases may occur as this method does not apply assumptions of probability theory and sampling errors [5]. Non-probability sampling allows for a less-restricted selection of participants, often removing the need for a sampling frame, such as a register of residents. However, this means that the probability of each case being chosen from a target population is not known (it may even be zero). There are five common non-probability sampling methods according to type of selection, likelihood of representativeness, extent of sample control, and overall cost and effort [6, 7]: • Quota: Sampling based on various quota variables, assuming their variability is the same as in the population • Purposive: Sampling by using personal judgement to select cases that will best enable to meet the research objectives • Snowball: Sampling by asking one respondent to establish contact with other potential (hard-to-reach) respondents • Self-Selection: Sampling by allowing each case to identify its interest to take part in the survey - often a crowdsourced task with mutual benefit (see also [8]) • Convenience: Sampling by haphazardly selecting those cases that are easiest to obtain Recruitment concept for an app-based travel-survey Developed by the Swedish company Trivector, TravelVu enables users to GPS-track their everyday movements and activities in a personal timeline. Learning algorithms suggest modes of transport and types of activity that can be traced in an interactive map. Travel characteristics can be adapted with several editing functions available through the app. For survey data, each travel day must be confirmed. Additionally, gamified pushmessages motivate users to correct their data. In doing so, participants receive a summary of their trips in distance and time. The desired sample target for this study was 1,000 individuals, selected from all relevant population groups within Dresden. From 14 Oct. to 24 Nov. 2019, anyone predominantly located in the city could participate. Due to legal regulations, respondents had to be 18 years or older. Potential participants were asked to answer an inapp questionnaire on socio-demographic attributes and to record their travel behaviour for at least seven days. The survey was called “Dresden in Bewegung” (Dresden in Motion), promoted with “Towards climatefriendly urban transport by app”, encouraging the contribution to a unique data basis for a better and more sustainable urban transport in the future. A non-probability sampling approach was used with a two-stage design, combining various sampling techniques: • First, haphazard sampling was used to reach the advised overall sample size, including mainly crowdsourcing and convenience sampling as well as techniques like snowball sampling. • Second, during the dynamic sampling phase, systematic selection was used to address specific, underrepresented groups and to minimise the risk of skewness. This was mainly a combination of purposive sampling and quota sampling as a comprehensive method for sample monitoring (age, gender, and post codes). As a part of the recruitment concept, resources were-accessed for spreading the survey via various communication channels. With aspects of participant motivation and technical support included, a broad-based and individually tailored recruitment concept was formed: Project Webpage: Available in German and English, this page provided information on how to participate as well as details on how to use the app and contact the technical support. Press Releases: In cooperation with the TU Dresden and the City of Dresden, two press releases were issued, to which several news media outlets responded (crowdsourcing). Social Media: A Facebook project page was set up to attract new participants as well as inform and motivate active ones through video posts explaining functions of the app, notices on the support, or posts on survey progress (crowdsourcing). These were shared by various persons, institutions, an action group (snowball sampling), and distributed throughout Facebook groups (convenience sampling). In the second survey stage, Facebook and Instagram ads were applied (purposive sampling) to target underrepresented groups regarding age and city districts. Local Media: A short promotional ad was shown on screens inside the Dresden tramcars, which was additionally broadcasted on an online TV channel (convenience sampling). Project Ambassadors: Two external supporters in particular contributed to the distribution (crowdsourcing): The City of Dresden spread the survey via its company mobility management e-mail list; the survey was also put on the Department of Transport Planning website. The TU Dresden utilised two student e-mail circulars and the monthly student newsletter to spread the survey. Both supporters also promoted the survey on their own social media channels. Printed Material: Posters were hung at specific points in the city, and brochures and post cards were displayed and handed out in the city centre as well as at the weekly market. The elderly were primarily targeted during this process (convenience & purposive sampling). In the second phase, these were additionally distributed to mailboxes along randomly selected routes in underrepresented districts (quota sampling and random route). Results During the course of the survey, there was a certain dropout of participants: 1,032 persons joined the survey, of which 941 answered the background questionnaire and 871 respondents recorded their trip information. By confirming travel days, a net sample of 618 participants contributed with travel behaviour data. With only 30 percent further dropout, those that remained collected data from seven days or more. This resulted in nearly 8,500 confirmed travel days, with an average of 13.7 days corrected by the participants, corresponding to a median of 10 days. INTERNATIONAL European Friedrich-List-Prize Internationales Verkehrswesen (72) 3 | 2020 60 For analysis, only the data from participants living inside the city (by post codes from the questionnaire) was used; this makes it possible to draw comparisons on register data. Figure 1 shows the distribution of age, gender as well as across the city’s districts: The sample consists of a higher share of young and a clearly low share of elderly people compared to the city’s register data from September 2019. However, users are more evenly distributed across the ten city districts. Regarding gender distribution, men are overrepresented to some extent. Participants also stated how they were made aware of the survey, which makes it possible to assess the recruitment process itself. The access modes most often mentioned were the TU Dresden e-mail circulars (42 %), followed by news (14 %), and Facebook (10 %), but also word-of-mouth advertising was referred to quite often (9 %). Social media ads and print distributions had observable precise and group-specific effects, especially through random routes. Platforms such as Instagram and Twitter as well as the ads shown in the tramcars were mentioned the least. When calculating the cost effectiveness of the recruitment, costs were related to the number of net participants (see figure 2). With reference to the net sample of n = 618, a cost of EUR 14.73 per participant was calculated. This is about a quarter less costly when compared to “Mobility in Cities - SrV” (a traditional cross-sectional household travel survey in Germany) with about EUR 20.50 per person in 2018. By correlating the net participants to specific access modes, it is apparent that ambassador-based recruitment was the most cost-effective and the distribution of printed material was the least of all. Discussion and Conclusion Some recruitment instruments such as the e-mail circulars showed very strong and specific survey success, though they also presented challenges in terms of addressing missing groups. Thus, effects to specific groups need to be considered, and better-performing alternatives should be tested in the future. The question remains if smartphone-based tracking, combined with non-probability sample recruitment, is a suitable survey method for elderly people - at least for Germany in 2020. In future applications, possible solutions could be, e.g., specifically confining the population to be addressed, providing additional “traditional” survey modes, or offering intensive support for elderly people during recruitment and data collection, eventually supplemented by a random sample. In conclusion, sample representativeness is a core quality criterion of traditional (household) travel surveys and a necessary prerequisite for making this data applicable for practical transport modelling and planning. This issue needs to be at least critically discussed and reflected towards the objectives of a travel survey. However, even if non-probability sampling strategies risk the potential for bias, they bring valuable advantages to recruitment in terms of flexibility, reactivity, and cost years 40% 50% 60% 50% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Dresden Sample Reference Population Male Female 8% 15% 5% 3% 2% 13% 4% 10% 26% 15% 11% 9% 9% 5% 6% 16% 7% 10% 10% 15% 0% 5% 10% 15% 20% 25% 30% 35% 37% 47% 12% 4% 10% 36% 22% 32% 0% 10% 20% 30% 40% 50% 60% 18-24 25-44 45-59 60+ Figure 1: Dresden net sample composition compared to register data from September 2019 Register Data: [9, 10] European Friedrich-List-Prize INTERNATIONAL Internationales Verkehrswesen (72) 3 | 2020 61 effectiveness, and may increase overall participant motivation — especially in combination with a user-friendly and appealing app design. As an area of future research, comparability of mobility patterns gathered by traditional (household) travel surveys and GPS-based data collections needs to be studied in more detail. ■ REFERENCES [1] Hubrich, S. and Wittwer, R. (2017): Effects of Improvements to Survey Methods on Data Quality and Precision - Methodological Insights into the 10th Wave of the Cross-Sectional Household Survey ”Mobility in Cities - SrV”. Transportation Research Procedia. Shanghai, pp. 2276-2286. https: / / doi.org/ 10.1016/ j.trpro.2017.05.436 [2] Verzosa, N.; Greaves, S. and Ellison, R. (2017): Smartphone-Based Travel Surveys: A Review. Working Paper. Retrieved from https: / / ses.library.usyd.edu.au/ handle/ 2123/ 19540 [3] He, S.; Wang, Z. and Leung, Y. (2017): Applying Mobile Phone Data to Travel Behaviour Research: A Literature Review. Travel Behaviour and Society. https: / / doi.org/ 10.1016/ j. tbs.2017.02.005 [4] Kuhnimhof, T.; Bradley, M. and Anderson, R.S. (2018): Workshop Synthesis: Making the Transition to New Methods for Travel Survey Sampling and Data Retrieval. Transportation Research Procedia, pp. 301-308. https: / / doi.org/ 10.1016/ j.trpro.2018.10.055 [5] Fowler, F.J. (2008): Survey Research Methods. 4th ed. Thousand Oaks: Sage Pubn Inc. ISBN: 978-1-4129-5841-7. Retrieved from https: / / dx.doi.org/ 10.4135/ 9781452230184 [6] Adams, K. and Brace, I. (2006): An Introduction to Market & Social Research: Planning & Using Research Tools & Techniques. London; Philadelphia: Kogan Page. ISBN: 978-0- 7494-4377-1 [7] Saunders, M.N.K.; Lewis, P. and Thornhill, A. (2016): Research Methods for Business Students. 5th ed. Essex: Pearson. ISBN: 978-1-292-01662-7 [8] Estellés-Arolas, E. and González-Ladrón-de-Guevara, F.G. (2012): Towards an Integrated Crowdsourcing Definition. Journal of Information Science. Vol. 38. https: / / doi. org/ 10.1177/ 0165551512437638 [9] City of Dresden (2019): City Population [Bevölkerungsbestand]. Retrieved from https: / / www.dresden.de/ media/ pdf/ statistik/ Statistik_1221_Lebensbaum_Quartal_Tabelle.pdf [10] City of Dresden (2019): Population across the City’s Districts (main residence) on 31 December 2019 [Bevölkerung am Ort der Hauptwohnung nach Stadtteilen am 31.12.2019]. Retrieved from www.dresden.de/ media/ pdf/ statistik/ Statistik_1219_Quartal_HW_n_ ST.pdf Johannes Weber, Dipl.-Ing. Research Associate, Chair of Integrated Transport Planning and Traffic Engineering, TU Dresden (DE) johannes.weber1@tu-dresden.de Stefan Hubrich, Dr.-Ing. Research Associate, Chair of Integrated Transport Planning and Traffic Engineering, TU Dresden (DE) stefan.hubrich@tu-dresden.de Rico Wittwer, PD Dr.-Ing. habil. Research Associate, Chair of Integrated Transport Planning and Traffic Engineering, TU Dresden (DE) rico.wittwer@tu-dresden.de Regine Gerike, Univ.-Prof. Dr.-Ing. Head of Chair, Chair of Integrated Transport Planning and Traffic Engineering, TU Dresden (DE) regine.gerike@tu-dresden.de Costs (including related working hours) Amount Subtotal Costs per Participant Netted by Access Mode Press Releases 161.00 € 1.63 € Writing and Coordinating with Ambassadors Social Media 1,640.00 € 22.47 € Facebook Page Setup, Creating Posts and Ads Ads in Local Media 380.00 € 76.00 € Public Transport TV Ambassadors 81.00 € 0.22 € Preparing Material for e.g. E-mail Circulars Printed Material 3,120.00 € 183.53 € Post Cards: Design, Display & Random-Route Distribution 4.000 1,300.00 € Brochures: Design, Display & Random-Route Distribution 2.000 1,150.00 € Posters: Design & display 100 670.00 € Project Webpage 661.00 € Setup and Maintenance Non-recruitment working hours translated to costs 3,060.00 € 10-week period (preparation phase: 4 weeks; data collection: 6 weeks) Total 9,105.00 € Total Costs per Net Participant n = 618 14.73 € Figure 2: Calculation of survey costs including respective working hours 18 th European Transport Congress (ETC) in-Rostock T he European Platform of Transport Sciences - EPTS Foundation e.V. - invites to the 18th European Transport Congress (ETC), which will be held in Rostock-Warnemünde, Germany, from 12-14 October 2020. The topic of the congress: “Innovative transport systems in European logistic networks - chances for non-metropolitan regions” All presentations will be held in English or German, with simultaneous translations. Contacts, conference program, and registration are available at www.epts.eu/ etc2020. Please register until 30.09.20 (online only). “We are looking forward to meeting you in Rostock”