Seismic Vulnerability and Risk Assessment: ML Approach

The IAStructE – IIIT Hyderabad student chapter of the Earthquake Engineering Research Centre organized an online ‘Guest Lecture’ on 29th August 2024. The lecture titled “Seismic Vulnerability and Risk Assessment: ML Approach” was delivered by Dr. Ratnesh Kumar, Professor, Department of Applied Mechanics, Visvesvaraya National Institute of Technology (VNIT), Nagpur. This event welcomed postgraduate students, research scholars, structural engineers, industry professionals, and faculty members.

The speaker began the lecture by signifying the need to predict the earthquake behaviour of the structures, to enlighten the local bodies and the government to reduce the loss of economy and human lives in the near future. He shared quite an interesting point that structures act as a shield to protect humans from natural hazards like floods, cyclones, etc., but the same acts as killers in the case of earthquakes. The talk proceeded with the concept of seismic vulnerability and the different methods for its risk assessment. The lecture focused on understanding the seismic risks associated with different regions and the vulnerabilities present in various building types. The use of Machine Learning (ML) and Geographic Information System (GIS) software for risk assessment was clearly illustrated through a case study in Nagpur city using the Socio-Economic Cluster approach. Other assessment-aiding parameters such as identifying the performance point using the demand and capacity curves and types of mapping were discussed further. The talk also shed light on the challenges faced in risk assessment and construction practices. The lecture concluded by highlighting the potential of combining various models using ML and GIS techniques for improved accuracy. In the end, an engaging question and answer session took place between the attendees and the speaker, where the insightful discussion and clarifications provided further enrichment of the participant’s understanding of the topic.

Visitors Count