Augmented asset management in railways – Issues and challenges in rolling stock

Abstract

Managing assets in railway, including infrastructure and rolling stock, efficiently and effectively is challenging. The emerging digital technologies and Artificial Intelligence (AI) are expected to augment the decision making in Asset Management (AM) and Fleet Management (FM). The AI technologies need to be adapted to the specific needs of any industrial domain, e.g. railways, to facilitate the implementation and achievement of the overall business goals. This adaptation is denoted as ‘Industrial AI’(IAI). IAI for railways infrastructure and rolling stock, is dependent on an appropriate technology roadmap reflecting necessary know-hows. The IAI roadmap aims to provide a strategic and executive plan to augment managing railway assets i.e. ‘Augmented Asset Management (AAM)’. AAM can be applied through an end-to-end secure platform for e.g. data sharing among stakeholders, the development of analytics, and model sharing through distributed computing. AAM in railways can be enhanced through implementation of a generic fleet management (FM) approach. In the FM approach, any population of assets with common characteristics and also the relationship of the asset to the fleet is considered. This paper aims to develop and propose a concept for AAM enabled through IAI and digital technologies to provide augmented decision support through a secure platform, for AM in railways. A FM approach towards a holistic operation and maintenance of assets, based on a System of Systems thinking, for AAM in railways is applied for population of infrastructure assets and rolling stock assets with common characteristics. Finally, a taxonomy of issues and challenges, in the application of AAM to FM in railways is provided. The data for this taxonomy has been collected from railway organizations through iterative rounds of interviews. This taxonomy can be used for research and development of frameworks, approaches, technologies, and methodologies for AAM in railways.

Optimization Simulation of Locomotive Semi-Active Suspension Control Based on Fuzzy Control and Detection of Electromechanical Equipment

“…The active suspension system can provide damping force proportional to the absolute motion speed of the object to be damped, so the adjustment principle of the damper in the semi-active suspension system is to make the damping force provided by the semi-active suspension system close to that provided by the active suspension system [11]. The damping force provided by the active suspension system (ideal damping force) is expressed as:…”

A framework for now-casting and forecasting in augmented asset management

“…The level of automation in AM can be enhanced by augmenting the decision making related to AM processes using AI tools and technologies. The concept of Augmented Asset Management(AAM) proposed by (Kumari et al 2021), is based on the provision of a platform that aims to facilitate integration and transformation of data into information, knowledge and context models that are based on available sources of data, keeping the domain specific challenges at its center, and adapting the analytics to the needs of the end-user. AAM is expected to increase the efficiency and effectiveness of AM through the utilisation of AI and digital technologies.…”

“…AM of complex technical system-of-systems needs crossorganizational operation and maintenance, asset data management and context-aware analytics (Kumari et al 2021). This creates the need for a common and distributed platform to integrate, analyze, and share information regarding overall asset health and performance.…”

“…It is challenging to model the behavior of each individual in a fleet as it increases the complexity of the architecture and infrastructure of the AI solution, with requirement for data acquisition and model development for every individual operating in different contexts. The Indi-vidual2Fleet and Fleet2Individual approach proposed by (Kumari et al 2021) can be used to simplify the infrastructure of the AI solution. The Fleet2Individual and Individu-al2Fleet approach is based on making generalisations for the fleet and identifying the individual behvaiors that are outliers from the fleet.…”

“…This has been introduced as a top down, i.e. Fleet2Individual, and bottom up, i.e., Individual2Fleet, approach to analytics (Kumari et al 2021).…”