23 November 2022
BIG DATA ASSISTS WITH MAINTAINANCE IN OPERATIONAL CONDITION
ENSURING EQUIPMENT AVAILABILITY AT ALL TIMES
Keeping Machine in Operational Condition (MOC)
MOC ensures the maintenance of equipment and integrates logistics: supplier management, procurement, storage and distribution of parts and technical elements: monitoring and processing of technical conditions.
Maintenance is broken down into various technical levels of intervention that is carried out at different levels of operational responsibility, depending on the sector.
Extracting information from unstructured data
In order to monitor maintenance activities, it is necessary to use intelligence, based on indicators from data in the Information System (IS). To define the indicators, the user must go through the following steps:
- Know the location and nature of data in the IS (including structured, unstructured, internal, external, versatile and permanent data)
- Define the expected performance
- Understand the variables that can influence and optimise the decision-making process
The data will then be fed into OPPSCIENCE’s solution. Once enhanced, it will be integrated into the MOC system via the Bee4sense platform, which contains three technologies: natural language processing, artificial intelligence and machine learning.
Key steps for data retrieval:
- Retrieval and analysis of all information, including intervention reports
- Extraction of named entities (products, components, equipment, organisations, people, etc.)
- Extraction of events (installations, incidents, maintenance) with their dates and locations
- Detection of anomalies and prevention of incidents or breakdowns
- Semantic reconciliation between incidents and good anticipation or processing practices
- Analysis and typing of event sequences via the machine learning solution
The Bee4sense platform gathers and correlates information accumulated from manufacturing, operations, and maintenance of equipment.
Optimising the decision-making process
This information is structured, but unstructured information can be added from messaging tools or projects such as reports, operational information from suppliers, subcontractors, operators and regulatory information, for example manufacturing or recycling constraints and information external to the company that is often difficult to consider.
Within weeks, users can have a single point of access to all relevant document information in MOC processes.
The semantic analysis module developed by OPPSCIENCE adds a layer of intelligence to the platform. All unstructured data is read and organised by named entities to provide users with an initial level of organisation. It is then compared to older data available in the systems or in Bee4sense to find links and correlations. Once the new organisation has been updated, users will have a ‘360°’ view of the data identified by them and the system. They will then follow the data at programmed frequencies and will have ‘graph’ views, allowing them to understand the links that unite the information in the form of named entities.
The intelligence that OPPSCIENCE’s IAM technologies bring to business issues such as MOC improve the data quality to ensure that the right decision is made.