The dynamics of a moving train are usually described by means of a motion model based on Newton's second law. This model uses as input track geometry data and train characteristics like mass, the parameters that model the running resistance, the maximum tractive effort and power, and the brake rates to be applied. It can reproduce and predict train dynamics accurately if the mentioned train characteristics are carefully calibrated. The model constitutes the core element of a broad variety of railway applications, from timetabling tools to Driver Advisory Systems and Automatic Train Operation. Among the existing train motion model calibration techniques, those that use operational data are of particular interest, as they benefit from on- board recorded data, capturing the train dynamics during operation. In this literature review article we provide an overview of the train motion model calibration techniques that have been published in the scientific literature between January 2000 and December 2021 and either use operational data or can be minimally adapted to use it. To this end, we present a critical overview of the existing train motion model calibration approaches, distinguishing online calibration that analyzes data on- the-go and offline calibration that analyzes historical data batchwise. We propose a research agenda and highlight some potential goals to be tackled in the near future: from devising accurate online calibrators for eco-driving applications to quantitizing the physical sources of parameter variation. Last, we discuss practical recommendations for practitioners and scholars inferred from the current state of the art.

@en

In this chapter, applications of artificial intelligence (AI) in railway traffic planning and management (RTPM) are discussed. To begin, a definition of AI is offered with a particular emphasis on its relationship with RTPM. This is followed by a systematic literature review of the state-of-the-art of AI in RTPM covering strategic, tactical, and operational challenges. Next, a transferability analysis is conducted of AI approaches for traffic planning and management from related sectors to railways, specifically from aviation and road transport. The results show that the majority of AI research in RTPM is still in its infancy. Several future research areas that are important to academic and professional communities in AI and RTPM are identified based on reviews and analysis of transferability.@en

This chapter presents applications, challenges, and opportunities for the integration of artificial intelligence in rail transport, based on the current results of the European project Roadmaps for AI integration in the rail sector (RAILS). Past and ongoing research directions are briefly outlined, and then the regulatory landscape is presented as well as the main barriers to overcome. Some technical aspects are addressed to provide some valuable references, and a high-level description of ongoing research work is given, spanning from innovative studies on smart maintenance, collision avoidance, delay prediction, and incident attribution analysis to visionary scenarios such as intelligent control and virtual coupling.@en

The COVID-19 pandemic has imposed a dramatic effect on the mobility habits of both passengers and freight in the rail sector. Since the relaxation of COVID-19 restrictions worldwide, rail transport has been revitalised gradually. However, the new normal emerges with unprecedented issues, such as changed travel behaviour, lost profits, and a lack of personnel. In this paper, we determine the arising challenges due to COVID-19 and pandemics in general and subsequently propose several solutions to tackle these challenges in rail transport. These solutions cover multidisciplinary aspects such as passenger demand management, freight demand management, service design, automation, decentralisation and advanced railway technologies. By reviewing the relevant literature on COVID-19, public transport and particularly rail transport, we synthesise and identify promising lines of research that should devote more attention to a more efficient, effective and sustainable rail transport service. This paper provides policymakers, researchers, railway infrastructure managers and undertakings with an overview and an outlook for the impacts of the pandemic crisis and similar situations. It supports decision-making with more evidence and facilitates rail transport to restore its performance and reach its societal goal.

@en

Disruptions occur frequently in railway networks, requiring timetable adjustments, while causing serious delays and cancellations. However, little is known about the performance dynamics during disruptions nor the extent to which the resilience curve applies in practice. This paper presents a data-driven quantification approach for an ex-post assessment of the resilience of railway networks. Using historical traffic realization data in the Netherlands, resilience curves are reconstructed using a new composite indicator, and quantified for a large set of single disruptions. The values of the resilience metrics are compared across disruptions of different causes using Welch's ANOVA and the Games-Howell test. Additionally, representative resilience curves for each disruption cause are determined. Results show a significant heterogeneity in the shape of the resilience curves, even within disruptions of the same cause. The proposed approach represents a useful decision support tool for practitioners to assess disruptions dynamics and propose best measures to improve resilience.

@en

Train movement dynamics are usually modelled by means of Newton's second law. The resulting dynamic equation can be very precise if the parameters that it depends on are determined accurately. However, these parameters may vary in time and show wide variations, making the calibration task nontrivial and jeopardizing the performance of a broad variety of applications in the railway industry: from timetable planning and railway traffic simulation to Driver Advisory Systems and Automatic Train Operation. In this article, the online train motion model calibration problem is addressed with a special focus on energy-efficient on-board applications. To this end, location and speed measurements are assumed to be available for a train running under normal operation conditions. A well-known real-time parameter estimation algorithm, the Unscented Kalman Filter, is combined with a driving regime calculator and a post-processing module in order to obtain bounds and statistics of parameters such as the maximum applied tractive effort and power, the applied brake rates, the cruise speed and the length of the final coasting and braking. The proposed framework is tested in a case study with real data from trains operating on the Eindhoven-’s-Hertogenbosch corridor in the Netherlands. Results obtained show that UKF is able to track the speed and location measurements and to estimate the parameters that model the running resistance in the dynamic equation. The proposed driving regime and the post-processing modules can determine the current regime accurately and give a deeper insight into the variations of the driving style, respectively.

@en

Trains consist of one or more railway vehicles called rolling stock, which need interior and exterior cleaning and small technical checks on a daily basis. These services are executed at service locations (SLs). Scheduling rolling stock servicing tasks during an operational day is important to guarantee the fulfilment of servicing deadlines. Public transport companies face large scheduling problems, especially those with 24-hour-a-day operation. The expected increase in transport frequencies enhances the need for improving scheduling servicing tasks during an operational day. Therefore, the Rolling Stock Servicing Scheduling Problem (RS-SSP) is modelled comprising a MILP model. Complying with the planned timetable, the RS-SSP maximises the RS units being serviced during daytime. The RS-SSP allows RS exchanges between RS units having completed servicing and operating RS units requiring servicing. Due to this RS Exchange Concept, the number of RS units visiting the SL during daytime can be increased. The proposed RS-SSP model has been tested on a real-life case from the Dutch railways. For multiple scenarios, the model was able to exchange all running RS. Consequently, the capacity usage at SLs can be increased by the RS-SSP by shifting some of the excessive workload to daytime, and thus solving the capacity shortages.

@en

The Netherlands have one of the highest rail electrification rates in the EU with over 75% of the railway network electrified (European Comission, 2018), offering environment-friendly trains operation. However, in order to achieve carbon neutral railway sector by 2050, significant investments are required to further improve environmental performance from trains operation, especially in regional nonelectrified networks with passenger services typically provided by diesel multiple unit (DMU) vehicles. Due to their low utilization, full electrification of such networks is often not economically viable, thus solutions are mainly sought in alternative propulsion system technologies, such as hydrogen fuel-cell multiple unit (FCMU) and battery-electric multiple unit (BEMU) vehicles (Klebsch et al., 2019). One of the main challenges in introducing BEMU trains is determining the electrification plan for the railway network, while satisfying requirements related to quality of service, maintaining current timetable, and vehicle-specific constraints. Previous research on BEMUs operation is mainly focused on continuous partial lines electrification, or eventually limited scenario analysis on intermittent electrification (Abdurahman et al., 2021), with the optimization-based methods still lacking in the literature. This study aims to fill this gap by proposing a method for developing an optimal electrification plan, while minimizing total costs and considering several electrification alternatives for each track section. @en

Nowadays it is widely accepted that Artificial Intelligence (AI) is significantly influencing a large number of domains, including railways. In this paper, we present a systematic literature review of the current state-of-the-art of AI in railway transport. In particular, we analysed and discussed papers from a holistic railway perspective, covering sub-domains such as maintenance and inspection, planning and management, safety and security, autonomous driving and control, revenue management, transport policy, and passenger mobility. This review makes an initial step towards shaping the role of AI in future railways and provides a summary of the current focuses of AI research connected to rail transport. We reviewed about 139 scientific papers covering the period from 2010 to December 2020. We found that the major research efforts have been put in AI for rail maintenance and inspection, while very limited or no research has been found on AI for rail transport policy and revenue management. The remaining sub-domains received mild to moderate attention. AI applications are promising and tend to act as a game-changer in tackling multiple railway challenges. However, at the moment, AI research in railways is still mostly at its early stages. Future research can be expected towards developing advanced combined AI applications (e.g. with optimization), using AI in decision making, dealing with uncertainty and tackling newly rising cybersecurity challenges.

@en

Demand for railway transportation keeps on growing. Therefore, a thorough understanding of the capacity of railway networks is crucial. In this paper, the well-known compression method based on max-plus algebra is extended. A number of challenges are addressed to apply this compression method to large and complex networks, such as the one considered in this paper. Some trains have to be split artificially, while keeping the parts together during the compression. The trains should also be ordered explicitly, since there is no part of the infrastructure used by all trains. The results in this paper indicate that it is possible to thoroughly analyse the capacity by the adjusted compression method for large and complex networks, but the results should be interpreted with care. The results show, for instance, that the capacity occupation heavily depends on the size of the network that is considered and that it is not easy to give a clear, practical interpretation of the capacity occupation. Nevertheless, the method allowed to determine a number of critical paths and, even more importantly, a number of critical resources in the zones considered.

@en

An accurate train motion model is a key component of a wide spectrum of railway applications, from timetabling algorithms to Automatic Train Operation systems. Therefore, model calibration has become crucial in the railway industry, although this topic has not received the attention and recognition in academia that its practical relevance deserves. Several data-driven techniques have been devised to calibrate train dynamics models, although an overview that describes the current state of the art in the field and highlights the following steps to be researched is still missing in the literature. Thus, this article has four main goals. First, giving a brief insight into the broad variety of techniques used for train motion model calibration, focusing on those techniques that use on-board measurements and are applicable in railway operation. Second, highlighting the main research steps to be tackled, considering the current main challenges in railway research. Third, outlining practical recommendations to practitioners who need to calibrate their algorithms and applications. And fourth, contributing to giving train motion model calibration its due recognition.@en

Employing regenerative braking in trains contributes to reducing the amount of energy used, especially when applied to commuter trains and to those used on very dense suburban networks. This paper presents a method to fine-tune the periodic timetable to improve the utilization of regenerative energy and to shave power peaks while maintaining the structure and robustness of the original timetable. First, a mixed-integer linear programming model based on the periodic event scheduling framework is proposed. A set of feasible timetables is determined and optimized with the aim of increasing synchronized acceleration and braking events at the same station, and maintaining the timetable robustness at the specified level. Next, a local search algorithm is developed to optimize the timetable such that the power peak value is minimized. The max-plus system model is adopted to estimate the delay propagation. Monte Carlo simulation is used to evaluate the utilization of regenerative energy and power peaks in random delayed circumstances. The proposed method was adopted to fine-tune the 2019 timetable for a sub-network of the Dutch railway. In the case of on- time scenarios, the optimized timetable increases the regenerative energy usage by almost 290% and decreases the 15-minute power peaks by 8.5%. In the case of delay scenarios, the optimized timetable outperforms the original timetable in terms of using regenerative energy and shaving power peaks.

@en

During railways operations, unplanned events might occur which can result in rail traffic being heavily impacted. The paper proposes a passenger-centred resilience assessment for disruption scenarios which consist of multiple simultaneous disruptions. It combines train traffic operations, passenger flows and network restoration. To evaluate resilience, an optimization-based approach has been developed for solving the new infrastructure restoration and transport management (IRTM) problem. Additionally, this approach develops mitigation plans for the best infrastructure restoration and traffic recovery and it captures the time-dependent transport network performance during disruptions. The approach is general with respect to types of disruptions, and can be applied for evaluation against short disruptions (1–2 h) as well as more substantial ones (multiple days or weeks). The performance of the proposed approach has been demonstrated on a Dutch railway network. Furthermore, the resilience of the system is assessed against the critical infrastructure disruption scenarios in the network. This optimization-based approach shall enable decision makers to quantify accurately impacts of multiple disruptions by considering the created inconveniences to passengers in the railway operation due to these disruptions.

@en

Highly utilized railway networks require regular infrastructure maintenance. Different track sections often need to be closed for entire days to carry out engineering works, which makes the regular timetables no longer feasible and thus adjusted railway services and temporary alternative services need to be planned. We introduce the Multimodal Alternative Services for Possessions (MASP) problem to support the planning of alternative services, from the passenger and transport operator points of view, including an adjusted train timetable, bus-bridging services and extra train services. The MASP problem is formulated based on the Service Network Design Problem and the Vehicle Routing Problem. To solve it efficiently, we develop a solution framework that incorporates heuristics based on the column and row generation with mixed-integer linear programming. The developed framework provides the optimized alternative service routes, schedules and passenger flows routing. We demonstrated the performance of the MASP solution framework on the real-life Dutch railway network. The results show that the MASP framework is capable of efficiently generating alternative services to route passenger flows affected by possessions with a very limited increase in the total passenger costs compared to a scenario with no link closures. High computational efficiency is observed even for highly disrupted networks.

@en

Due to the covid19 crisis, public transport (PT) systems are facing new challenges. Regarding restrictive measures such as physical distancing and the successive returning of passengers after the “intelligent lockdown”, significant lack of transport capacity can be expected. In this paper, the transport capacity of a PT network is assessed, using a mathematical passenger route choice and train scheduling model. By analysing the overall number of transported passengers and the resulting link and train utilization; the networks capabilities of facilitating different demands under capacity restrictions (e.g. physical distancing) are addressed. The analysis is performed on the Dutch railway network. The results show that at most 50% of the pre-covid19 demand can be transported, while most of the trains will be highly utilized reaching their maximum occupation. Thus, significantly more parts of the network are becoming highly utilized, leading to a more congested and vulnerable system than in normal conditions before covid19.

@en

The aim of this paper is to review existing publicly available and open artificial intelligence (AI) oriented datasets in different domains and subdomains of the railway sector. The contribution of this paper is an overview of AI-oriented railway data published under Creative Commons (CC) or any other copyright type that entails public availability and freedom of use. These data are of great value for open research and publications related to the application of AI in the railway sector. This paper includes insights on the public railway data: we distinguish different subdomains, including maintenance and inspection, traffic planning and management, safety and security and type of data including numerical, string, image and other. The datasets reviewed cover the last three decades, from January 1990 to January 2021. The study revealed that the number of open datasets is very small in comparison with the available literature related to AI applications in the railway industry. Another shortcoming is the lack of documentation and metadata on public datasets, including information related to missing data, collection schemes and other limitations. This study also presents quantitative data, such as the number of available open datasets divided by railway application, type of data and year of publication. This review also reveals that there are openly available APIs—maintained by government organizations and train operating companies (TOCs)—that can be of great use for data harvesting and can facilitate the creation of large public datasets. These data are usually well-curated real-time data that can greatly contribute to the accuracy of AI models. Furthermore, we conclude that the extension of AI applications in the railway sector merits a centralized hub for publicly available datasets and open APIs.@en

Artificial Intelligence (AI) is becoming pervasive in most engineering domains, and railway transport is no exception. However, due to the plethora of different new terms and meanings associated with them, there is a risk that railway practitioners, as several other categories, will get lost in those ambiguities and fuzzy boundaries, and hence fail to catch the real opportunities and potential of machine learning, artificial vision, and big data analytics, just to name a few of the most promising approaches connected to AI. The scope of this paper is to introduce the basic concepts and possible applications of AI to railway academics and practitioners. To that aim, this paper presents a structured taxonomy to guide researchers and practitioners to understand AI techniques, research fields, disciplines, and applications, both in general terms and in close connection with railway applications such as autonomous driving, maintenance, and traffic management. The important aspects of ethics and explainability of AI in railways are also introduced. The connection between AI concepts and railway subdomains has been supported by relevant research addressing existing and planned applications in order to provide some pointers to promising directions.

@en

Due to increasing railway use, the capacity at railway yards and maintenance locations is becoming limiting to accommodate existing rolling stock. To reduce capacity issues at maintenance locations during nighttime, railway undertakings consider performing more daytime maintenance, but the choice at which locations personnel needs to be stationed for daytime maintenance is not straightforward. Among other things, it depends on the planned rolling stock circulation and the maintenance activities that need to be performed. This paper presents the Maintenance Location Choice Problem (MLCP) and provides a Mixed Integer Linear Programming model for this problem. The model demonstrates that for a representative rolling stock circulation from The Netherlands Railways a substantial amount of maintenance activities can be performed during daytime. Also, it is shown that the location choice delivered by the model is robust under various time horizons and rolling stock circulations. Moreover, the running time for optimizing the model is considered acceptable for planning purposes.@en

In big cities, the metro lines usually face great pressure caused by huge passengers demand, especially during peak hours. When disruptions occur, passengers accumulate quickly at stations. It is of great importance for dispatchers to take passenger flow control into consideration for the traffic management to ensure passengers' safety and to maintain their satisfaction. This paper proposes an integrated disruption management model, which incorporates train rescheduling and passenger flow control. In this model, the train services can be short-turned, cancelled and rerouted, while the number of passengers entering a station is managed by controlling the station gates with consideration of the capacities of platforms and trains. Moreover, the number of passengers arriving at a station is calculated according to the origin-destination matrices. The objectives are to recover the train operation to the original timetable as soon as possible and to minimize the waiting time of passengers outside the stations. With the interaction between train services, passengers and station gates, an iterative metaheuristic approach is proposed to solve the integrated disruption management problem. Based on the data of a Beijing metro line, numerical experiments are conducted to test the proposed algorithm. The results demonstrate the importance of integrated disruption management and the effectiveness of our solution method.

@en

The performance and behaviour of critical infrastructure in case of disruptions is an important topic and we are still lacking of insights. Due to disruptions, infrastructure becomes unavailable and may force the trains and passengers to adapt. In this paper, we introduce a problem of railway network vulnerability from the perspective of passenger flows and train operations. We propose a new Railway Network Vulnerability Model (RNVM) to assess the vulnerability of the system by finding the critical combination of links, which cause the most adverse consequences to passengers and trains. To solve this challenging problem, we present a RNVM framework, which combines two heuristics based on column and row generation with mixed integer linear programming, to efficiently model alternative passenger flows and infrastructure constraints. The developed framework provides the critical combination of links, the corresponding passenger flows, train routes and timetables. We demonstrate the performance of the RNVM framework on the real-world instance of a part of the Dutch railway network. The results show that the RNVM framework can efficiently reassign passenger flows and reroute trains during disruptions. The results also reveal that the critical links are highly demand dependent rather than a static feature of the networks topology. Finally, the computation times remain small when increasing the number of disrupted links as well as the size of the passenger demand, which allows fast and efficient network vulnerability assessment.

@en