PostDoc Project
Title:
Formalization of knowledge applied to the domain of predictive maintenance for sustainable operations management
Dates:
2022/07/03 - 2024/07/02
Supervisor(s):
Description:
Sustainability issues are now at the heart of political and academic debates, as well as the central role of industrial organisations
as key stakeholders in global sustainable development. Therefore, sustainable manufacturing is a necessary paradigm for the
long-term survival of the manufacturing industry, which needs to prioritise sustainable value creation, in order to face several
challenges, such as the depletion of physical resources, stricter laws and regulations, and customer demand for better product
quality (Eslami, Dassisti, Lezoche, & Panetto, 2019).
Longer equipment life cycles and higher, sustainable performance can be achieved through effective maintenance, making this
business function fundamental to sustainable manufacturing. Maintenance processes have non-negligible impacts on the
sustainability objectives of manufacturing companies that should be studied to contribute to the scientific discussion as well as
to raise awareness among industrial actors about the relationship between maintenance and sustainability, and the strategic role
that the maintenance department can have (Franciosi, Voisin, Miranda, Riemma, & Iung, 2020). Therefore, maintenance
decisions should be made not only on the basis of technical and economic criteria, but by considering in a synergistic way the
impact of maintenance activities on the economic, environmental and social dimensions of sustainability.
Furthermore, a mandatory condition for sustainable operations management is to facilitate the decision-making of the
organisation's processes by relying on collaboration and the exchange and sharing of information (Eslami, Ashouri, & Lezoche,
2021). It is therefore necessary to ensure that the information exchanged has the same meaning for all processes involved
(Szejka, Canciglieri, Panetto, Rocha Loures, & Aubry), which leads to semantic interoperability. Semantic interoperability
allows to capture a shared (semantic) meaning from shared information based on the formalisation of process knowledge.
The use of an ontological engineering approach to formalise knowledge in maintenance and enable decision making in
maintenance has been discussed in the literature, and maintenance and asset management are among the business processes that
can benefit from the application of ontologies (Polenghi, Roda, Macchi, Pozzetti, & Panetto, 2021) : Indeed, ontology-based
approaches allow knowledge and data to be capitalised and could then help decision makers to identify the causes of abnormal
operations (Medina-Oliva, Voisin, Monnin, & Leger, 2014). Moreover, in the current and future industrial scenario of the smart
factory, i.e. knowledge-intensive, ontologies allow (Polenghi, Roda, Macchi, & Pozzetti, 2021):
1. to retrieve and properly distribute data/information among stakeholders for cross-functional decision making
2. Provide a single, standardised vocabulary, avoiding misinterpretation between stakeholders;
3. increase the content of information through their reasoning and inference capabilities.
The sustainable value chain can be created through the potential offered by the fourth industrial revolution often referred to as
Industry 4.0 (Stock, Obenaus, Kunz, & Kohl, 2018), which through its technologies enables the collection and elaboration of a
large amount of data from the different equipment in factories (Panetto, Iung, Ivanov, Weichhart, & Wang, 2019). For example,
the data generated by industrial assets, i.e. Cyber Physical Systems, being contextualised, generate information, then a potential
source of knowledge that must be extracted, formalised, and then potentially reused (Lezoche M., HDR dissertation, 2021).
Therefore, the large amount of data collected by these systems contains valuable information about industrial processes, which,
by applying data-driven approaches or machine learning methods, it is possible to find interpretative results for strategic
predictive maintenance decision making (Carvalho, Soares, Vita, Francisco, Basto, & Alcalá, 2019). For all these reasons, by
combining data collection through I4.0 technologies, artificial intelligence techniques and knowledge formalisation, it is
possible to achieve predictive maintenance strategies for sustainable smart manufacturing.
Consistent with the context of the project, what we propose in this study is to acquire knowledge about
- Ontological engineering methods aiming at developing an adequate ontology-based approach adapted to industrial
applications;
- Asset health management and predictive maintenance strategies;
- The impact of maintenance on the sustainable performance (i.e. economic, environmental and social) of manufacturing
companies.
The research questions (RQs) that this project aims to answer are :
- RQ1: How is knowledge modelled in the field of predictive maintenance?
- RQ2: What ontologies/knowledge maps exist in the field of predictive maintenance in the context of manufacturing and are
sustainability factors/indicators taken into account to make sustainable decisions in production processes?
- RQ3: What are the relationships between maintenance processes, asset management practices and their impacts on the
economic, environmental and social dimensions of sustainability?
- RQ4: How can these relationships be taken into account in existing ontologies?
- RQ5: How can an ontology engineering method be properly adopted to develop a knowledge-enriched maintenance approach
for sustainable operations management?
as key stakeholders in global sustainable development. Therefore, sustainable manufacturing is a necessary paradigm for the
long-term survival of the manufacturing industry, which needs to prioritise sustainable value creation, in order to face several
challenges, such as the depletion of physical resources, stricter laws and regulations, and customer demand for better product
quality (Eslami, Dassisti, Lezoche, & Panetto, 2019).
Longer equipment life cycles and higher, sustainable performance can be achieved through effective maintenance, making this
business function fundamental to sustainable manufacturing. Maintenance processes have non-negligible impacts on the
sustainability objectives of manufacturing companies that should be studied to contribute to the scientific discussion as well as
to raise awareness among industrial actors about the relationship between maintenance and sustainability, and the strategic role
that the maintenance department can have (Franciosi, Voisin, Miranda, Riemma, & Iung, 2020). Therefore, maintenance
decisions should be made not only on the basis of technical and economic criteria, but by considering in a synergistic way the
impact of maintenance activities on the economic, environmental and social dimensions of sustainability.
Furthermore, a mandatory condition for sustainable operations management is to facilitate the decision-making of the
organisation's processes by relying on collaboration and the exchange and sharing of information (Eslami, Ashouri, & Lezoche,
2021). It is therefore necessary to ensure that the information exchanged has the same meaning for all processes involved
(Szejka, Canciglieri, Panetto, Rocha Loures, & Aubry), which leads to semantic interoperability. Semantic interoperability
allows to capture a shared (semantic) meaning from shared information based on the formalisation of process knowledge.
The use of an ontological engineering approach to formalise knowledge in maintenance and enable decision making in
maintenance has been discussed in the literature, and maintenance and asset management are among the business processes that
can benefit from the application of ontologies (Polenghi, Roda, Macchi, Pozzetti, & Panetto, 2021) : Indeed, ontology-based
approaches allow knowledge and data to be capitalised and could then help decision makers to identify the causes of abnormal
operations (Medina-Oliva, Voisin, Monnin, & Leger, 2014). Moreover, in the current and future industrial scenario of the smart
factory, i.e. knowledge-intensive, ontologies allow (Polenghi, Roda, Macchi, & Pozzetti, 2021):
1. to retrieve and properly distribute data/information among stakeholders for cross-functional decision making
2. Provide a single, standardised vocabulary, avoiding misinterpretation between stakeholders;
3. increase the content of information through their reasoning and inference capabilities.
The sustainable value chain can be created through the potential offered by the fourth industrial revolution often referred to as
Industry 4.0 (Stock, Obenaus, Kunz, & Kohl, 2018), which through its technologies enables the collection and elaboration of a
large amount of data from the different equipment in factories (Panetto, Iung, Ivanov, Weichhart, & Wang, 2019). For example,
the data generated by industrial assets, i.e. Cyber Physical Systems, being contextualised, generate information, then a potential
source of knowledge that must be extracted, formalised, and then potentially reused (Lezoche M., HDR dissertation, 2021).
Therefore, the large amount of data collected by these systems contains valuable information about industrial processes, which,
by applying data-driven approaches or machine learning methods, it is possible to find interpretative results for strategic
predictive maintenance decision making (Carvalho, Soares, Vita, Francisco, Basto, & Alcalá, 2019). For all these reasons, by
combining data collection through I4.0 technologies, artificial intelligence techniques and knowledge formalisation, it is
possible to achieve predictive maintenance strategies for sustainable smart manufacturing.
Consistent with the context of the project, what we propose in this study is to acquire knowledge about
- Ontological engineering methods aiming at developing an adequate ontology-based approach adapted to industrial
applications;
- Asset health management and predictive maintenance strategies;
- The impact of maintenance on the sustainable performance (i.e. economic, environmental and social) of manufacturing
companies.
The research questions (RQs) that this project aims to answer are :
- RQ1: How is knowledge modelled in the field of predictive maintenance?
- RQ2: What ontologies/knowledge maps exist in the field of predictive maintenance in the context of manufacturing and are
sustainability factors/indicators taken into account to make sustainable decisions in production processes?
- RQ3: What are the relationships between maintenance processes, asset management practices and their impacts on the
economic, environmental and social dimensions of sustainability?
- RQ4: How can these relationships be taken into account in existing ontologies?
- RQ5: How can an ontology engineering method be properly adopted to develop a knowledge-enriched maintenance approach
for sustainable operations management?
Keywords:
Knowledge formalisation, ontology engineering, predictive maintenance, sustainable manufacturing
Conditions:
The duration of the contract is 24 months
The employer is the CNRS and the place of employment is the CRAN laboratory for 80% of the time and the company Vosges
Immobilier Entreprendre for the remaining 20%.
The salary will be 2100 euros net per month.
The expected profile is of a researcher who knows and practices the following research fields
- Ontology Engineering
- Knowledge Formalisation
- Formal analysis of concepts
- Maintenance engineering and management
- Predictive maintenance
- Sustainable manufacturing and circular economy
- Industry 4.0 and smart technologies
- Artificial intelligence techniques applied to industrial scenarios
- Digital twin and cyber-physical systems
- Human factors and human reliability in manufacturing systems
The employer is the CNRS and the place of employment is the CRAN laboratory for 80% of the time and the company Vosges
Immobilier Entreprendre for the remaining 20%.
The salary will be 2100 euros net per month.
The expected profile is of a researcher who knows and practices the following research fields
- Ontology Engineering
- Knowledge Formalisation
- Formal analysis of concepts
- Maintenance engineering and management
- Predictive maintenance
- Sustainable manufacturing and circular economy
- Industry 4.0 and smart technologies
- Artificial intelligence techniques applied to industrial scenarios
- Digital twin and cyber-physical systems
- Human factors and human reliability in manufacturing systems
Department(s):
| Eco-Technic systems engineering |
Funds:
The funding comes from the France Relance scheme
Publications:
