Siddharth Tallur

Indian Institute of Technology Bombay

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

Sauvik Banerjee

Indian Institute of Technology Bombay

Sunil A. Bhave

Purdue University West Lafayette

Ruchira Nandeshwar

Indian Institute of Technology Bombay

Pallabi Das

Indian Institute of Technology Bombay

Sheetal Patil

Indian Institute of Technology Bombay

Jeslin Thalapil

Indian Institute of Technology Bombay

Kiran Kondabagil

Indian Institute of Technology Bombay

Durgesh Tamhane

Indian Institute of Technology Bombay

Avani Kulkarni

Indian Institute of Technology Bombay

Shruti Sawant

Indian Institute of Technology Bombay

Cooperative Institutions

Indian Institute of Technology Bombay

119

Tata Institute of Fundamental Research

Physical Research Laboratory

Indian Institute of Science Bangalore

Cornell University

Inter-University Centre for Astronomy and Astrophysics

Savitribai Phule Pune University

California Institute of Technology

Raman Research Institute

University of Mumbai

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

Phase sensitive detection of extent of corrosion in steel reinforcing bars using eddy currents

arXiv (Cornell University) (2020)

Indrani Mukherjee Jinit Patil Sauvik Banerjee Siddharth Tallur

Corrosion of steel bars in reinforced cement concrete (RCC) structures leads to premature deterioration and increase in life cycle maintenance costs. Non-destructive testing (NDT) of incipient corrosion has been an impending task in this domain. We present a low cost sensing platform based on eddy current detection using anisotropic magnetoresistive (AMR) sensor to measure the extent of corrosion in steel reinforcing bars (rebars). The scheme employs phase-sensitive detection, wherein the phase shift in the sensor output (proportional to surface conductivity of the rebar) relative to the reference stimulus is measured by a lock-in amplifier and used to distinguish corroded and non-corroded rebars. The proof-of-concept sensor demonstration is able to resolve varying extents of rebar corrosion and can find potential applications as an NDT tool in a variety of industries.

Rebar

10.48550/arxiv.2001.03756

Cite

Citations (1)

Beyond 5 GHz excitation of a ZnO-based high-overtone bulk acoustic resonator on SiC substrate

arXiv (Cornell University) (2023)

Padmalochan Panda Soumyadip Chatterjee Siddharth Tallur Apurba Laha

This work describes the fabrication and characterization of an Au/ZnO/Pt-based high-overtone bulk acoustic resonator (HBAR) on SiC substrates. We evaluate its microwave characteristics comparing with Si substrates for micro-electromechanical applications. Dielectric magnetron sputtering and an electron beam evaporator are employed to develop highly c-axis-oriented ZnO films and metal electrodes. The crystal structure and surface morphology of post-growth layers have been characterized using X-ray diffraction (XRD), atomic force microscopy (AFM), and scanning electron microscopy (SEM) techniques. HBAR on SiC substrate results in multiple longitudinal bulk acoustic wave resonances up to 7 GHz, with the strongest excited resonances emerging at 5.25 GHz. The value of f.Q (Resonance frequency * Quality factor) parameter obtained using a novel Q approach method for HBAR on SiC substrate is 4.1 * 10^13 Hz which, to the best of our knowledge, is the highest among all reported values for specified ZnO-based devices.

Overtone

Crystal (programming language)

10.48550/arxiv.2304.04286

Cite

Citations (0)

Deep transfer learning strategy for efficient domain generalisation in machine fault diagnosis

Research Square (Research Square) (2022)

Supriya Asutkar Siddharth Tallur

Abstract Automated fault diagnosis algorithms based on vibration sensor recordings play an important role in determining the state of health of the machines. Data-driven approaches require a large number of supervised and labelled samples to build reliable models. The performance of such lab-trained models when deployed in practical use cases largely depends on their domain generalisation capability towards distinct distribution target domain datasets. These challenges are addressed to some extent by conventional parameter-based transfer learning techniques that fine-tune deeper (dense) layer parameters of the pre-trained networks when lower (convolutional) layers can capture more domain-specific information concerning the changing target domain distribution. In this work, we present a novel deep transfer learning strategy that utilises fine-tuning the parameters of convolutional layers and transfer of knowledge in the dense layers from the source domain for efficient domain generalisation and fault classification. The performance of this strategy is evaluated with two different target domain data sets, by studying the sensitivity of fine-tuning individual layers in the networks using time-frequency representations of the vibration signals (scalograms) as inputs. We observe that this strategy yields near-perfect accuracy, even for use cases where low-precision sensors are used for data collection, and unlabelled run-to-failure data with a limited number of training samples.

Transfer of learning

10.21203/rs.3.rs-2215201/v1

Cite

Citations (1)

A Low Power Impulse Radio Design for Body-Area-Networks

IEEE Transactions on Circuits and Systems I Regular Papers (2011)

Rajeev Dokania Xiao Y. Wang Siddharth Tallur Alyssa Apsel

This paper presents a low power radio design tailored to the short distance, low data rate application of body area networks. In our analysis we consider a comparison between traditional continuous wave radios and ultra wide band impulse radios for this application space. We analyze the energy/bit requirement for each of the architectures and discuss how a duty-cycled radio is better suited to low data rate applications due to practical design considerations. As a proof-of-concept we present the design and measured results of a duty-cycled, noncoherent impulse radio transceiver. The designed transceiver was measured to consume only 19 μW at a data-rate of 100 kbps. The design gives a BER of 10 ^-5 and works for a range of 2.5 m at an average Rx-sensitivity of -81 dBm. The designed transceiver enables both OOK and BPSK schemes and can be configured to use a pseudocoherent self-correlated signature detection and generation mechanism. This added functionality helps distinguish different types of pulses such as timing and data-pulses in real time. The transceiver was designed in a 90 nm CMOS process and occupies 2.3 mm ² area.

Transceiver

Duty cycle

Body area network

10.1109/tcsi.2011.2131250

Cite

Citations (56)

Dipstick-based pathogen detection for wastewater surveillance: Variability analysis using gage repeatability and reproducibility

medRxiv (Cold Spring Harbor Laboratory) (2024)

Shruti Ahuja Avani Kulkarni Kiran Kondabagil Siddharth Tallur

COVID-19 redefined the outlook on pandemic preparedness, accelerating research toward establishing a global consortium for wastewater surveillance. Due to sample heterogeneity and low pathogen loads, microbial concentration remains a key challenge in developing low-cost, point-of-use wastewater monitoring assays. To address this challenge, we have developed a simplified version of dipstick method for RNA capture and isolation from sub-milliliter sample volumes, which simplifies RNA isolation. Given the manual steps involved in executing the dipstick method, variability is a major concern. In this work, we assessed dipstick variability through a multi-operator gage repeatability & reproducibility (gage R&R) study. We focused on detecting pepper mild mottle virus (PMMoV) and bacteriophage Phi6 in wastewater samples collected from a sewage pumping station at IIT Bombay. Our study demonstrated that the repeatability and reproducibility for the dipstick method are less than the acceptability limit of 30%, and could detect changes in PMMoV load associated with change in the population density due to summer break in our campus. Phi6 is a widely accepted surrogate for enveloped viruses such as SARS-CoV-2, and was therefore also chosen to demonstrate utility of this method for monitoring spread of infectious diseases. This work underscores the effectiveness of gage R&R in assessing and understanding the sources of variations in such assays, including operator-induced and part-to-part differences, essential for developing robust, manually-operated assays.

Repeatability

10.1101/2024.11.07.24316947

Cite

Citations (0)

A Silicon Electromechanical Photodetector

Nano Letters (2013)

Siddharth Tallur Sunil A. Bhave

Optomechanical systems have enabled wide-band optical frequency conversion and multichannel all-optical radio frequency amplification. Realization of an on-chip silicon communication platform is limited by photodetectors needed to convert optical information to electrical signals for further signal processing. In this paper we present a coupled silicon microresonator, which converts near-IR optical intensity modulation at 174.2 MHz and 1.198 GHz into motional electrical current. This device emulates a photodetector which detects intensity modulation of continuous wave laser light in the full-width-at-half-maximum bandwidth of the mechanical resonance. The resonant principle of operation eliminates dark current challenges associated with convetional photodetectors. While the results presented here constitute a purely classical demonstration, the device can also potentially be extended to the quantum regime to realize a photon-phonon translator.

10.1021/nl400980u

Cite

Citations (19)

Non-destructive evaluation of a composite honeycomb sandwich panel with in-plane waviness damage using ultrasonic guided waves

Nondestructive Testing And Evaluation (2024)

Ramana Raja Buddala Siddharth Tallur Sauvik Banerjee

Composite honeycomb sandwich structures (CHSS) are often manufactured using an autoclave co-cure process, which can introduce in-plane waviness (IW) damage due to fibre misalignment. Such fibre waviness, whether in-plane or out-of-plane, significantly weakens the strength of these structures. Therefore, developing reliable non-destructive evaluation (NDE) techniques is essential for detecting IW. This study presents an NDE technique utilising ultrasonic guided waves (GW) to identify IW damage in CHSS panels. A numerical model was created using COMSOL Multiphysics, where IW damage is simulated through curvilinear coordinate physics, followed by a time-dependent study of GW propagation. This model simulates local fibre orientation changes due to IW damage, enhancing our understanding of GW interaction with IW. Experimental investigations are performed using contact-type transducers to corroborate the numerical observations. The presence of in-plane waviness causes a reduction in the group velocity of the A0 mode, a characteristic that can be exploited for detecting in-plane fibre waviness in CHSS. Also, parametric studies were performed to examine the effects of IW severity (aspect ratio) and increased IW area on GW characteristics. Finally, a technique for IW damage localisation is demonstrated using a convex hull algorithm based on changes in correlation coefficient values of the A0 mode.

Waviness

Honeycomb

Honeycomb structure

Sandwich panel

10.1080/10589759.2024.2363983

Cite

Citations (0)

A Reconfigurable Hardware Emulator of MEMS Gyroscopes with Built-in Error Source Models

2022 Joint Conference of the European Frequency and Time Forum and IEEE International Frequency Control Symposium (EFTF/IFCS) (2021)

Sanjoli Narang Siddharth Tallur

The paper presents a hardware emulator of MEMS vibratory gyroscopes that can be used for accelerating the characterization and verification of control electronics, by decoupling these tasks from the MEMS design and fabrication cycles. The hardware emulator is synthesized on a field-programmable gate array (FPGA). In addition to modeling the Coriolis effect, the emulator also models typical error sources in MEMS gyroscopes, namely quadrature error, Duffing non-linearity, thermomechanical noise etc. Preliminary results based on a prototype with user-controllable device parameters synthesized on a Xilinx Zynq-7000 FPGA (Digilent Zybo Z7) are presented in this paper.

Decoupling (probability)

10.1109/eftf/ifcs52194.2021.9604312

Cite

Citations (0)

Low-Cost Vibration Sensor for Condition-Based Monitoring Manufactured From Polyurethane Foam

IEEE Sensors Letters (2017)

Reshma Ajith Amit Tewari Dipti Gupta Siddharth Tallur

Ubiquitous vibration sensing forms a core requirement of Internet of Things (IoT) applications in condition-based monitoring (CbM). Such sensors can enable cost savings by identifying incipient failures in industrial machinery and, thereby, optimized maintenance schedule planning. Conventional piezoelectric and microelectromechanical systems (MEMS)-based vibration sensors developed for such applications cost upwards of tens and hundreds of dollars, limiting the scale of their deployment. In this article, we present an extremely inexpensive vibration sensor prepared with commercially available polyurethane foam that is commonly used for packaging of fragile goods. We present a process to coat the pores of the foam with conductive carbon ink to impart piezo-resistive properties to the material. A proof of concept realization of vibration sensor with 80-Hz sensing bandwidth is presented, along with experimental data demonstrating classification of vibration signals for different machine operating conditions. The spectral content of the measured vibration signal shows good agreement with spectral content of the audio recordings of corresponding acoustic measurements.

Resistive touchscreen

10.1109/lsens.2017.2773652

Cite

Citations (19)

Low-Cost Passive pH Sensor Fabricated on Scotch™ Tape

IEEE Sensors Letters (2020)

Ayan Majumder Subrat Kumar Pradhan Kasturi Saha Siddharth Tallur

We report the fabrication and characterization results of a simple and low-cost pH sensor fabricated using a graphite pencil to define a working electrode and silver paste to define a reference electrode on scotch tape. The sensor operation is based on potentiometric measurement and thereby insensitive to fabrication variations in shape of the electrode unlike amperometric and chemiresistive measurement techniques. The substrate of the disposable sensor is prepared by pasting scotch tape atop a piece of chart paper, and two types of sensors fabricated with 6B and 2B graphite pencils are tested with three solutions with different pH values. The sensor functions as a passive sensing tag without requiring any external power or stimulus, and the measured sensitivities of the pH sensors fabricated using 2B and 6B pencil carbon electrodes (PCEs) are $-4.54mV/pH$ and $-4.09mV/pH$ respectively.

Pencil (optics)

Amperometry

Potentiometric sensor

10.1109/lsens.2020.2982469

Cite

Citations (5)