eMaintenance [LAB] A Platform for Effective and Efficient Decision Making in Maintenance
Maintenance decision support and knowledge discovery form an important part of the decision support system of the organization to create additional value. Relevant information through knowledge discovery – as a part of the decision support system – is required by the decision makers to facilitate maintenance decision support.
Effective maintenance decisions with multiple shareholders, depend largely on information and data to estimate the Remaining Useful Life (RUL) of equipment. The decision support process needs a trusted DSS (Decision Support System) based on a knowledge discovery process, which is essentially consists of data acquisition, transition, fusion, analysis and information extraction and visualization (Figure 1).
Data Fusion
Data fusion is a prerequisite in the decision process to support management with data from heterogeneous sources or from multiple sensors. Knowledge discovery, when applied to maintenance decision support, uses eMaintenance concept to integrate the data mining and knowledge discovery. However, development of eMaintenance solutions for industrial applications faces a number of challenges which can be divided into organizational, architectural, infra-structural, content & contextual and integration categories.
Organizational challenges
Organisational challenges are mainly aspects related to enterprise resource management like the restructuring of the organizations involved in eMaintenance, resource planning (material, spare-parts), information and knowledge management and management of heterogeneous organizations.
Architectural challenges
Architectural challenges deal with development related to the overall architecture of eMaintenance solutions, for example framework for eMaintenance, models for decentralized data processing and analysis, model- based prognostic tools etc.
Infra-structural challenges
Are challenges related to the provision of necessary technologies and tools required when services are developed, implemented and managed in an enterprise. Some examples are network infra-structure, authentication and authorization of services and users, safety and security mechanisms etc.
Content and contextual challenges
Content and contextual challenges are mainly related to data and information provided through the eMaintenance services, for example appropriate ontology through which data from data sources can smoothly and seamlessly be integrated, quality assurance mechanisms ensuring the required data quality, mechanisms to manage uncertainty in data sets etc.
Integration challenges
Integration challenges are caused by issues around the coordination, orchestration, and integration of services and data managed by the eMaintenance solution. These are related to the management, interaction and interactivity of services, configuration management of eMaintenance services and enablement of integration capability across a multi-platform and technologies.
Figure 1. A generic knowledge discovery process.
Figure 2. The concept of eMaintenance Cloud.
Figure 3. Fact-based maintenance decision-making.
Establishment of knowledge discovery mechanisms for maintenance decision support can be facilitated through development of eMaintenance solutions through the eMaintenance cloud (Figure 2). An eMaintenance Lab facilitates knowledge discovery in maintenance using eMaintenance solutions. The development of these solutions as support to maintenance decision-making can be facilitated through the provision of a meta-level model, through which a range of concepts, models, techniques and methodologies can either be clarified and/or integrated.
eMaintenance Solutions
The rapid growth of information in servicebased economies has inspired a desire for a deeper understanding of information services. New information technologies have resulted in service platforms that offer a wide range of information services including user-generated and freely exchanged content in organized communities. Stateof- the-art research has focused on optimizing information services by modelling them as primarily based on transactions. However, it is a systematically understudied issue how to design information services for effective feedback loops in critical processes such as the process industry or transportation logistics.
The deployment and use of such complex technical systems are common in society and in industry. Many of the complex technical systems also have stringent requirements on safety, dependability and cost, which necessitate frequent updates and modifications in response to new developments in technology and changing functional requirements. Hence, correct situation and context-adaptation and timely information and information support are crucial for management, operation and maintenance to improve both the systems and the services that these systems are expected to provide. This can be facilitated by the utilisation of new and innovative Information & Communication Technology (ICT) manifested in emerging approaches such as eMaintenance. eMaintenance Cloud (eMC) materialise a set of inter-operable, independent and loosely coupled information services with its inherent infra-structure (Figure 2). This provides enterprise-wide, continuous, coordinated service support and service delivery functions for business, operation and maintenance processes.
Recently, eMaintenance targeted information services have received great attention in many industrial domains that range from various manufacturing industries to aviation, mining, transportation, shipping, process control and nuclear power industries. There are many incentives for pushing the eMaintenance agenda, which can be identified from three perspectives that strongly influence each other: business, societal and national.
From a business perspective, eMaintenance leads to the better understanding and management of problems and their underlying causes and thereby to more proactive, cost-effective and efficient maintenance. This is done by taking into consideration the context, the user (Gould & Lewis, 1985), the technology and the organization (Reason, 1990). The system is conducted to reduce the chance for error often caused by hidden problems in these types of complex systems (Reason, 1997). In turn this leads to lower Life Support Costs (LSC), improved capability to estimate Remaining Useful Life (RUL), prolonged lifecycle, increased sustainability of complex systems, improved availability performance and reduced Life Cycle Costs (LCC).
From a societal perspective, eMaintenance mainly implies a shift towards service- oriented and solution-oriented industry and better working environment. Access to the right information in a timely manner and its timely delivery to the right stakeholders strongly influence productivity and service provision.
However, eMaintenance aims to support fact-based maintenance decision-making (Figure 3) through establishment of an effective and efficient information logistics.
eMaintenance is a critical performance driver for enhanced competitiveness – for companies, regions and nations. It is very important to provide cost-effective maintenance services for these systems during their whole lifecycle. It is not the quality of the development results, but rather the quality of the maintenance and service work that will make the national economies take a substantial leap into the future success.
eMaintenance LAB
It is challenging to develop effective and efficient information logistics solutions for factbased maintenance decision support through eMaintenance:
Business Challenges are Asset Management, Performance Based Logistics (BPL), Contracted Logistic Support (CLS), sustainability, Key Performance Indicators (KPI), Lifecycle management.
Methodological Challenges like RAMS analysis, Risk management, Predict-and- Prevent instead of Fail-and-Fix, Prognostic Health Management (PHM), Data mining, Data analysis and Life Cycle Cost and Risk estimation.
Technological Challenges such as Asset monitoring, Cloud computing, Decentralised data processing; Embedded systems, Real- time & on-line data processing, Data integration, Ontology, Model development, Data acquisition, Context sensing, Content management, Service-orientation and Event-driven.
Today there are existing contributions (standards, tools, technologies, and methodologies) that deal with different aspects of the aforementioned challenges. However, there is still a huge need to develop an extensible mechanism (i.e. eMaintenance Cloud, eMC) through which holistic and innovative eMaintenance solutions can be designed, developed and established (Figure 4).
Figure 4. A schematic diagram of eMaintenance LAB’s functionality.
However, the development of an eMaintenance Cloud and its inherent components brings new challenges for industry and academia. To overcome these challenges, the Division of Operation and Maintenance Engineering at Luleå University of Technology has established, in close collaboration with industrial partners from various branches, an eMaintenance Cloud materialised through eMaintenance LAB. The aim of eMaintenance LAB is to provide artefacts that facilitate development of fact-based innovative decision support solutions for maintenance. The LAB supports a multi-purpose crossdomain usage scenario for research, education and innovation in industry and academia (Figure 4).
The eMaintenance LAB at Luleå University of Technology (LTU) facilitates the operation of eMC through a set of autonomous information services. The services are demonstrated as applications (apps) and provided via an infra-structure, the Industrial Apps Pool.
Today, two context-adapted clouds are under development, one aimed for commercial and the other for mining railways. Data from various data-sources such as forces data, wheel profile data etc. is fused and integrated in real-time and used for advanced data analysis of asset health and for decision support in maintenance planning and execution. It is expected that in the future the eMaintenance Lab will be the hub of maintenance- related research at Luleå University.
Most of the services provided by eMaintenance LAB are available through the internet. Currently, these services are hosted at two locations in the northern part of Sweden, one in Luleå and the other one in Kiruna.
