top of page

Korkut

Kaynardag

Google

Scholar

Linkedin

Research 

Gate

Education

Education

The University of Texas at Austin (2016 to present)

Ph.D. in Civil Engineering

Bogazici University (2013 to 2016)

M.Sc. in Civil Engineering

Bogazici University (2008-2013)

B.Sc. in Civil Engineering

Czech Technical University (2012 Fall)

Exchange semester 

Technical Interests and Skill

Experience

  • Lead Data and System Engineer in Graduate Research Assistant Rile- The University of Texas at Austin (2016 to present)

  • To detect defects (outliers) in railway tracks at higher speeds, designed and developed a non-contact laser sensor-based predictive maintenance system framework using vibration and acoustic data.

  • Decreased cost by 40% and increased speed by 25%.

  • Developed noise reduction and data cleaning algorithms: supervised learning, artificial neural networks (semi-organizing maps), time-series methods, wavelets, signal smoothing algorithms, and bilinear interpolation for noise signal detection/noise-free signal estimation.

  • Applied hypothesis testing and multivariate statistical methods combined with transfer function-based system identification methods to enhance anomaly detection and defect classification.

  • Integrated an LSTM autoencoder deep-learning anomaly detection approach with feature selection (TensorFlow) to the framework.

  • Parametrically tested different signal processing and system identification methods to improve noise reduction.

  • Developed numeric computational models (acoustic wave propagation simulations), and integrated them with experimental data to simulate the developed predictive maintenance system.

  • Collaborated to develop analytical transfer function-based methods to compute the attenuation and propagation zones of acoustic wave modes in continuous periodically supported solid (i.e., rail).

  • Carried out novel speckle noise tests for Laser Doppler vibrometers.

  • Carried out laboratory and field testing of the system.

  • Prepared 11 journals (5 published, 6 in submission), 2 of which are on deep learning, gave 5 conference presentations and 1 invited talk.

  • Data and System Engineer in Graduate Research Assistant Role - Bogazici University (2013 to 2016)

  • Applied noise reduction, signal processing, and system identification methods to vibration signals for condition monitoring and predictive maintenance of a tall building, a wind turbine, a suspension bridge (1560 meters long), a masonry building, and 13 masonry bridges.

  • Applied convex-optimization-based model updating methods to enhance the model predictability of engineering structures’ simulations as well as probabilistic methods for reliability estimation.

  • Carried out dynamic and seismic assessments through analytical methods as well as FEM simulations.

  • Developed a real-time remote sensing and automated dashboard.

  • Published 1 journal and made 4 conference presentations.

Additional Machine Learning Projects

  • Vehicle price prediction: Applied linear regression models with different regularization approaches as well as XGBoost regression on Python to predict vehicle prices.  

  • Bank account fraud detection: highly imbalanced dataset, created a pipeline using supervised models (KNN, support vector machines, random forest, ADA boost, and XGBoost) and improved detection performance using stacking and up/down sampling on Python.

  • Customer segmentation: Applied unsupervised classification (i.e., DBScan, MeanShift, Ward, K-means) and dimensionality reduction to perform customer segmentation analysis on Python.

Courses - Certificates -Testing Experience - Soft Skills

Courses

Advanced Mathematics                                               Structural Dynamic               

Earthquake Engineering                                             Finite Element Method I & II   

Structural Reliability                                                   Structural Health Monitoring

Advanced Mechanics of Materials                            Advanced Behavior of Concrete

Advanced Steel Behavior                                            Acoustic I & II                 

Stochastic Process-Estimation- Control                Machine Learning (Audited)     

Neural Networks   (Audited)

Certificates

University of Michigan Phyton Course – Coursera (May 2022)

2 certificates in Python programming and data structures

IBM Machine Learning Course – Coursera (April 2023)

4 certificates in Explotary Data Analysis and Feature engineering, Regression, Supervised and Unsupervised learning

International Courses

1.  4th Summer School on Smart Materials and Structures, 2015, Trento, Italy 

2. COST-Action: Quantifying the value of structural health monitoring 2nd workshop, 2015, Istanbul, Turkey 

3.  Workshop on Bridge Health Monitoring, 2015, Istanbul, Turkey

Sensor and Data Acquisition System Experience

Laser Doppler vibrometry            

Acoustic Sensors

Oscilloscope                    

Linear potential meters

Strain gauges                 

Modal mass shakers

Accelerometers (force-balance and piezoelectric)

DAQ systems (PXI, Dewesoft, Kinemetrics)

Sensor integration of accelerometers, acoustic sensors, and lasers.

Soft Skills

Communication and data storytelling: prepared 12 peer-reviewed research papers; prepared 18 technical reports for funding agencies; made 8 conference presentations; supported funding proposals; got 2 commercialization and entrepreneurship awards; and made pitches to funding agencies.

Teamwork and leadership: created and led 9 different research and technical groups.

Problem-solving and critical thinking: developed algorithms and solutions to 7 complex data-based predictive maintenance and structural health monitoring/NDT research projects.

Education and collaboration: created educational sections on engineering systems and their dynamics as well as signal processing, optimization, and machine learning on the personal webpage; mentored students; teaching assistant to several engineering courses; peer-reviewer for engineering journals.

Time management: helped plan the scheduling of different data-based predictive maintenance and structural health monitoring/NDT projects; successfully completed the projects on time; carried out several tests in laboratories and on fields on time.

Publications and Presentations

Journal Publications

 

  1. M Shamszadeh, K Kaynardag, C Yang, S Salamone “Deep learning and system identification integrated detection of rail defects using a non-contact rail defect detection system”, IEEE Transactions on Automation Science and Engineering, (in preparation).

  2. K Kaynardag, A Pirotta, S Salamone “Using scanning measurements to represent the speckle noise in rail measurements performed by an laser Doppler vibrometer placed on a moving platform”, Measurement, (to be submitted in summer 2023).

  3. C Yang, K Kaynardag, S Salamone “LSTM autoencoder based automated railway anomaly detection using laser Doppler vibrometer measurements”, Structural Health Monitoring, (submitted).

  4. C Yang, K Kaynardag, S Salamone “Missing rail fastener detection based on laser doppler vibrometer measurements”, Journal of Nondestructive Evaluation, (submitted).

  5. C Yang, K Kaynardag, S Salamone “Investigation of wave propagation and attenuation in periodic supported rails using wave finite element method”, Acta Mechanica, (available online).

  6. K Kaynardag, C Yang, S Salamone “A rail defect detection system based on laser doppler vibrometer measurements”, NDT & E International, (accepted).

  7. K Kaynardag, C Yang, S Salamone “An impulsive noise filter for rail vibration measurements performed through a laser doppler vibrometer placed on a moving platform”, Mechanical Systems and Signal Processing, (in revision).

  8. C Yang, K Kaynardag, S Salamone “Evaluation of fastening modeling approaches for dynamic assessment of rail based on finite element method”, Journal of Engineering Mechanics, 148 (9), 04022048, 2022

  9. K Kaynardag, C Yang, S Salamone “The numerical simulations to examine the interaction of train-induced guided waves with transverse cracks”, Transportation Research Record, (available online), 2022.

  10. K Kaynardag, G Battaglia, A Ebrahimkhanlou, A Pirotta, S Salamone “Identification of bending modes of vibration in rails by a laser doppler vibrometer on a moving platform”, Experimental Techniques, 45 (1), 13-24, 2021

  11. K Kaynardag, G Battaglia, C Yang, S Salamone “Experimental investigation of the modal response of a rail span during and after wheel passage”, Transportation Research Record, 2674 (12), 15-24, 2020

  12. K Kaynardag, S Soyoz “Effect of identification on seismic performance assessment of a tall building”, Bulletin of Earthquake Engineering, 15 (8), 3227-3243, 2017

Conference Proceedings

  1. C Yang, K Kaynardag, S Salamone “LSTM autoencoder for anomaly detection in rails using laser doppler vibrometer measurements”, 14th  International Workshop on Structural Health Monitoring (IWSHM), 2023 (submitted)

  2. S Soyoz, E Karcioglu, E Aytulun, K Kaynardag, S C Pevlan, A Karadeniz, “Dynamic identification - model updating – seismic performance assessment of stone arch bridges”, Proceedings of the 4th Conference on Smart Monitoring, Assessment and Rehabilitation of Civil Structures, 2017

  3. S Soyoz, U Dikmen, N Apaydin, K Kaynardag, E Aytulun, S Senkardesler, N Catbas, H Lus, E Safak, M Erdik, “System identification of bogazici suspension bridge under hanger replacement”, Procediea Engineering, 199, 1026-1031, Part of X International Conference on Structural Dynamics, EURODYN 2017, 2017.

  4. H Sesigur, G Erol, S Soyoz, K Kaynardag, S Gonen, “Repair and retrofit of ketchaouo mosque in Algeria”, In Structural Analysis of Historical Constructions: Anamnesis, Diagnosis, Theraphy, Controls, CRC Press, 1824-1831, 2016

  5. K Kaynardag, S Soyoz, “Seismic performance assessment of a tall building based on real-time monitoring”, SECED 2015 Conference: Earthquake Risk and Engineering towards a Resilient World, 2015

  6. K Kaynardag, S Soyoz, “Structural health monitoring of a tall building”, Second European Conference on Earthquake Engineering and Seismology, 2014

Conference Presentations

  1. K Kaynardag, C Yang, S Salamone “Detection of rail defects using guided ultrasonic waves and laser doppler vibrometry”, Society for Experimental Mechanics Conference-XLI, Austin, 2023

  2. K Kaynardag, C Yang, S Salamone “Rail defect detection by noncontact vibration measurements”, Engineering Mechanics Institute Conference, John Hopkins, Baltimore, 2022

  3. K Kaynardag, G Battaglia, S Salamone “Applicability of laser Doppler vibrometer placed on a moving platform for rail vibration measurements”, Transportation Research Board Annual Meeting 2021, Washington DC, 2021

  4. K Kaynardag, A Ebrahimkhanlou, S Salamone “Modal based detection of cracks in railway tracks and applicability of moving laser Doppler vibrometer”, Engineering Mechanics Institute Conference, MIT, Boston, 2018

Invited Talk

  1. “Development of a noncontact laser Doppler vibrometer based rail damage detection system”, Weekly Acoustic Seminar at Mechanical Engineering Department, The University of Texas at Austin, Austin, 2023

Attended Conferences

  1. The AI Summit and IoT World Austin, Austin, TX, USA, November 2022

bottom of page