| Project ID |
BITS-SRIP/AB3FAB/2026 |
| Project Title |
Machine-learning driven analysis for engineering plasmonic sensors or Simulation-driven Heterostructure-based Infrared Explosive Label-free Detection (SHIELD) |
| Project Description |
Hybrid structures are providing /coming up with enhanced performance parameters. How to optimise these structures, is a challenging task. A high-performance SPR sensor with an enhanced Figure of Merit (FOM) will be designed using a data-driven optimization approach. The developed dataset will incorporate multiple operating wavelengths (900, 1100, and 1550 nm), enabling improved generalisation and robust model behaviour. Once trained, the ML models will assist in cross-verifying sensor outputs and effectively identify and remove outliers, thereby enhancing the reliability of the design process. The work will focus on expanding the dataset to further enhance prediction
accuracy, minimise errors, and strengthen statistical performance indicators, such as R². Overall, this study will demonstrate that data-driven methods can deliver fast,
accurate, and physically meaningful optimisation of SPR sensors while avoiding labour-intensivee trial-and-error approaches.
or This project is aimed at computational designs and optimization of a Mid-IR plasmonic sensing platform using 2D graphene–hBN heterostructures and engineered plasmonic metasurfaces for label-free detection of explosives and hazardous chemicals. With our advanced electromagnetic simulations (FEM) and material modelling (DFT), we will engineer configurations for high surface-enhanced infrared absorption (SEIRA), spectral selectivity, and detection sensitivity. Mid-IR plasmonic sensors are sensitive to molecular vibrational modes associated with explosives and harmful chemicals, but deployed measurements suffer from limited sensitivity and a non-specific nature. The utilization of 2D graphene-hBN heterostructures provides a unique and significant opportunity because of graphenes tunable surface plasmons in Mid-IR and hBNs low-loss phonon-polaritonic capabilities to confine light at the nanometres scale [1,2]. Combined with engineered metasurfaces, they can offer enhanced electromagnetic fields leading to observable SEIRA phenomena - a crucial requirement for detecting hazardous analytes at trace levels.
Institutions in India (IITs, IISc) and DRDO-supported labs have begun to use electromagnetic simulation to design plasmonic heterostructures for environmental and security sensing applications. However, the main remaining gap in this field is that most projects are limited to conceptual studies or basic prototyping, and nearly never does a project that uses simulation-driven work to systematically match plasmonic resonances with explosive vibrational fingerprints, while also considering practicable device challenges that include fabrication limitations and environmental noise. SHIELD aims to help close this gap by incorporating the latest modeling, electromagnetic design, and computational optimization capabilities to produce field-ready mid-IR sensors. By prioritizing simulation-driven approaches in this context, SHIELD can help close the gap from learning an idea on the bench to deploying a manufactured technology system while increasing national security and environmental monitoring capabilities. |
| Project Discipline |
ECE, EEE, CSE, PHY, Electronics and computer |
| Faculty Name |
Pankaj Arora |
| Department |
Department of Electrical & Electronics Engineering |
