Project
MetA HAZOP – Methodology for Automated HAZOP Studies
Development of an AI-supported Methodology for conducting automated or semi-automated HAZOP Studies
dashboard Objectives
As part of the research project, a comprehensive methodology for conducting automated or partially automated HAZOP studies is being developed. With the help of knowledge representations and inference engines, the steps of hazard identification, risk assessment and risk-based definition of safeguards are carried out.
The knowledge representations are to be created or expanded on the basis of existing HAZOP studies. Using artificial intelligence methods such as natural language processing systems, HAZOP tables and other necessary documents are to be correctly recorded and semantically evaluated so that they can then be converted into knowledge representations using suitable methods such as ontology learning.
AI-supported HAZOPs will not only save time and costs in the future, but above all make risk analyses more objective – for greater safety in processes and plants.
Figure: Screening of HAZOPs using Natural Language Processing as a basis for the AI-supported creation of ontologies.
dashboard MILESTONES
Concept for automated HAZOP studies
Digitization of existing HAZOPs through natural language processing
Automated generation of knowledge representations
Structuring of ontology modules as well as test and validation methodology
Input objects for automated HAZOPs
dashboard Overview
Automated HAZOPs
Natural Language Processing
Ontology Learning
dashboard Motivation
Technical risk analyses are carried out in industry in order to systematically identify the hazards of technical systems for people and the environment, assess possible consequences and define suitable safeguards. In practice, risk analyses, such as HAZOP studies, are carried out by a team of experts at great expense of time. The risk posed by a plant must be regularly reviewed during its life cycle and may need to be reassessed several times in case of modifications or changed boundary conditions. The results of the risk analyses are significantly influenced by the composition of the team and the experience of the participants.
The time, personnel and costs involved can be significantly reduced by using automated concepts to carry out risk analyses. In addition to supporting the teams, automated concepts can help to expand and harmonize the knowledge base and reduce subjective assessments and evaluations. Documentation processes can also be automated.
dashboard PUBLICATIONS
Publications in preparation.
PROJECT MANAGER:
STAFF | CONTACT:
person Marius Baechle, M.Sc.
email Send email
phone +49 721 6699 4780
FORSCHUNGSSCHWERPUNKTE:
More Projects
SISProof – Safety Instrumented System Proof
Development of soft sensors for SIS monitoring and evaluation of limiting factors.
sRMC – Safety related Remote Process Monitoring & Control
Applying AI in acoustic-based sensor systems for gas leakage detection
ZEBrA – Zero-Emission Breathing Application for Storage Tanks
Development of an emission-free storage tank with construction and testing of a prototype.
Cover photo source "Tube Geometry", RiIM project: Harald Hoyer from Schwerin, Germany, CC BY-SA 2.0, via Wikimedia Commons
.jpg){kind=link}
Cover photo source "Winstainforth", SafeDDT project, CC BY-SA 3.0, via Wikimedia Commons
