Valeria Vittorini
Please Note
9 records found
1
This paper provides an overview of the main results achieved within the Horizon 2020 Shift2Rail project named RAILS (Roadmaps for Artificial Intelligence Integration in the Rail Sector). The RAILS roadmapping process provided state-of-the-art, taxonomy, future research directions, and recommendations in three macro areas: Railway Safety and Automation, Predictive Maintenance and Defect Detection, and Traffic Planning and Management. RAILS findings shed light on the potential of intelligent technologies and provided essential guidelines for integrating machine learning into next-generation smart railways.
Moving Block (MB) and Virtual Coupling (VC) rail signalling will change current train operation paradigm by migrating vital equipment from trackside to onboard to reduce train separation and maintenance costs. Their actual deployment is however constrained by the industry's need to identify configurations of MB and VC signalling equipment which can effectively guarantee safe train movements even under degraded operational conditions involving component faults. In this paper, we analyse the effectivity of MB and VC in safely supervising train separation under nominal and degraded conditions by using an innovative approach which combines Fault Tree Analysis (FTA) and Stochastic Activity Networks (SAN). An FTA model of unsafe train movement is defined for both MB and VC capturing functional interactions and cause-effect relations among the different signalling components. The FTA is used as a basis to apportion signalling component failure rates needed to feed the SAN model. Effective MB and VC train supervision is analysed by means of SAN-based simulations in the specific scenario of an error in the Train Position Report (TPR) for five rail market segments featuring different traffic characteristics, namely high-speed, mainline, regional, urban and freight. Results show that the thresholds of the design variables depend on the considered signalling system alternative and the investigated market segment. In particular, the TPR delay threshold allowed for MB is higher than for VC. This means that to ensure a safe train movement, VC cannot absorb a TPR delay of longer than 1.5 s, which corresponds to the mainline market segment. For MB instead, the results show that the maximum TPR delay can reach 3.9 s for high-speed and freight railways. In addition, results showed that the integration of an FTA in a SAN model can provide a better understanding of the safety-performance behaviour of a system where VC showed a higher number of braking indications with respect to MB for the same TPR error failure rate. This means that for VC to effectively supervise the train separation at the same safety level as MB, we would need to have a much higher reliability of the TPR. The overall approach can support infrastructure managers, railway undertakings, and rail signalling suppliers in investigating the effectiveness of MB and VC to safely supervise train movements in scenarios involving different types of degraded conditions and failure events. The proposed method can hence support the railway industry in identifying effective and safe design configurations of next-generation rail signalling systems.
Railway systems are complex given their interconnectivity with sub-systems wherein each contains multiple components. Virtual Coupling (VC) is a next-generation railway signalling technology that advances Moving Block (MB), also known as European Train Control System Level 3 (ETCS L3). Some pilot implementations exist for MB. However, VC is still a visionary system and involves several safety issues due to the relative braking distance between trains. Therefore, it is important to evaluate the safety of this system to understand whether it is feasible for deployment. This paper performs a preliminary safety and reliability study by introducing a fault tree (FT) model to investigate the possible causes that lead to an unsafe train movement for MB and VC. To this aim, a FT model is initially built for the MB system, considering the system configurations and interactions between wireless devices, onboard and trackside equipment. Then, the FT model of the VC system is derived on top of the one for MB and the differences are highlighted between the FT elements of the two systems.
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.
Deliverable D4.1
Real-Time Traffic Rescheduling Algorithms and Perturbation Management and Hazard Prevention in Moving-Block Operations
A non-vital early-warning prediction model of hazardous MB traffic conditions is also proposed which includes a short- and a medium-term hazard identification method. In the short-term, potentially hazardous MB traffic condition are identified as violations of safety-critical threshold values of design variables relating to MB train operations (e.g. driving reaction times), the GNSS system (e.g. GNSS error or latency) and/or the GSM-R layer (e.g. MA communication delay). Safety-critical thresholds of the different design variables are identified by means of an extensive sensitivity analysis which uses a Stochastic Activity Network built for MB within WP2. In the medium-term warnings of potentially hazardous MB conditions are instead triggered whenever RECIFE-MILP detects track occupation conflicts in geographical areas with limited GNSS and/ or GSM-R signal availability, such as deep valleys or tunnels. The defined models contribute to the definition of an optimised automated Traffic Management System for Moving Block which can also support traffic dispatchers in preventively avoiding the occurrence of potentially dangerous MB traffic conditions. ...
A non-vital early-warning prediction model of hazardous MB traffic conditions is also proposed which includes a short- and a medium-term hazard identification method. In the short-term, potentially hazardous MB traffic condition are identified as violations of safety-critical threshold values of design variables relating to MB train operations (e.g. driving reaction times), the GNSS system (e.g. GNSS error or latency) and/or the GSM-R layer (e.g. MA communication delay). Safety-critical thresholds of the different design variables are identified by means of an extensive sensitivity analysis which uses a Stochastic Activity Network built for MB within WP2. In the medium-term warnings of potentially hazardous MB conditions are instead triggered whenever RECIFE-MILP detects track occupation conflicts in geographical areas with limited GNSS and/ or GSM-R signal availability, such as deep valleys or tunnels. The defined models contribute to the definition of an optimised automated Traffic Management System for Moving Block which can also support traffic dispatchers in preventively avoiding the occurrence of potentially dangerous MB traffic conditions.
Artificial Intelligence in Railway Transport
Taxonomy, Regulations and Applications
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.