Deep Learning and Machine Learning for Vault and Temple Structure Analysis: Shri Padmanabhaswamy Temple
Author : Arya A Deshmukh, Shivam Y Shingare, Sunil T Rathod, Deepvijay V Chavan and Dr. Aparna Vaidyanathan
Abstract :
The Shri Padmanabhaswamy Temple in Kerala, India, holds one of the world's largest treasures — estimated at over USD 22 billion — locked inside seven underground vault chambers. Most vaults are accessible, but Vault B has never been opened due to religious beliefs. This paper explains, in simple technical terms, how modern Artificial Intelligence (AI) techniques — specifically Deep Learning (DL) and Machine Learning (ML) — can be used to study the temple's vaults and its 1,000-year-old stone structure without physically disturbing them. We describe six AI tools:
- Image recognition models to identify and catalogue artefacts in the open vaults;
- Ground-Penetrating Radar with AI to see inside Vault B through walls;
- D laser scanning with AI to map the entire temple structure;
- Sensor-based monitoring with AI to detect damage in stone walls;
- AI prediction models to forecast how the building will age; and
- AI image generation to digitally restore damaged artefacts.
Each method is explained with its purpose, how it works, and what results it can deliver. This research shows that AI can help protect and study this national heritage site without breaking any religious rules or legal restrictions.
Keywords :
Deep Learning, Machine Learning, CNN, YOLOv8, GPR, LiDAR, LSTM, Vault Analysis, Temple Structure, Heritage Conservation, Padmanabhaswamy.
