SURFACE
Sara Khoshnevisan
Trying to make better engineering decisions when the ground beneath us is uncertain, using reliability-based design, machine learning, nondestructive testing, and remote sensing for infrastructure resilience.
BIO
Who I am.
My research has a clear thread, even though I did not plan it that way at the start.
My doctoral work at Clemson focused on robust geotechnical design: keeping designs reliable when soil properties are uncertain. When I joined UC, I carried that reliability mindset into new applications: machine learning for predicting soil properties, Bayesian modeling of in-situ test correlations, and probabilistic liquefaction assessment.
Over time, the data-driven work raised a new question: what if the biggest limitation is the data itself? That is what pulled me into nondestructive testing and intelligent compaction, where the focus shifts toward generating better field measurements rather than extracting more from limited ones. Even then, continuous high-quality data is rarely possible in practice, which raises a parallel question: where should we sample, and how do we predict what lies between the points we actually test?
I did not map this progression out from the beginning. It developed through experience, student-driven questions, and the practical needs of the agencies supporting the research. The result is a clear arc, from managing uncertainty in design to improving the field measurements that designs depend on.
THRUSTS
Four lines of investigation.
My program spans four interconnected thrusts, unified by one theme: managing uncertainty in geotechnical design and characterization.
Reliability-based & robust geotechnical design
The foundation of my program, beginning with my doctoral work on efficient robust design: developing methods that stay reliable even when soil properties are uncertain. Includes LRFD calibration, robust optimization, and design under spatial variability.
Liquefaction assessment & seismic geotechnics
Probabilistic liquefaction assessment built on long-standing international collaborations, now applying XGBoost with Bayesian hyperparameter optimization and genetic programming to derive interpretable, closed-form susceptibility models, alongside work on geological uncertainty in hazard characterization.
Machine learning & data-driven geotechnics
Machine learning and Bayesian modeling for geotechnical problems: CPT-based soil classification, resilient modulus prediction, CPT-SPT correlations using hierarchical Bayesian models, and interpretable prediction. The emphasis throughout is on models that are transparent and trustworthy for practicing engineers.
Nondestructive testing & intelligent compaction
The largest and fastest-growing part of my program, with one overarching goal: cost-effective, continuous nondestructive testing of soil and other geomaterials. Two FHWA/INDOT projects develop acoustic-wave and accelerometer-based techniques that infer soil stiffness in real time from roller operation, while a UC internal grant extends the approach to image analysis and deep learning for compaction assessment. Together they move from proof of concept toward practical DOT guidelines.
RECORD
Recovered cores from the record.
A book chapter, 28 journal articles (3 under review), 28 peer-reviewed conference papers, and a technical report. Selected cores are logged below in reverse stratigraphic order; the complete record lives in the CV.
| Core ID | Yr. | Title & Authors | Venue | Status |
|---|---|---|---|---|
| CORE-01 | 2026 |
Practical Intelligent Compaction with Modest Accelerometers: Real-Time Soil Stiffness via Roller Accelerometer Vibration Analysis
Revised version submitted May 2026 · Originally TRB, moved to TRR
|
Transportation Research Record | Accepted w/ Rev. |
| CORE-02 | 2026 |
Within-Group Data Leakage and Inflated R² in Geotechnical ML: A Resilient Modulus Study
Submitted May 11, 2026
|
J. Transportation Eng., Part B: Pavements | Under Review |
| CORE-03 | 2026 |
Audio-Based Non-Destructive Soil Compaction Monitoring Using Training-Free Wavelet Scattering Features
Originally submitted Jan 27, 2026 · Major revisions in progress
|
Journal of GeoData and AI | Accepted w/ Major Rev. |
| CORE-04 | 2026 |
Enhancing 3D Soil Characterization through Machine Learning from CPT Data
|
GeoCongress 2026 | Published |
| CORE-05 | 2025 |
Region-Specific CPT-SPT Correlations for Cohesionless Soils: A Hierarchical Bayesian Approach
|
Geotech. & Geological Eng. · Q1 | Published |
| CORE-06 | 2024 |
LiDAR-Based 3D Litho-Stratigraphic Models Calibrated with Limited Boreholes
|
Engineering Geology · Q1 | Published |
| CORE-07 | 2024 |
Use of Machine Learning Methods to Obtain a Reliable Predictive Model for Resilient Modulus of Subgrade Soil
|
Purdue JTRP · Tech. Report | Published |
| CORE-08 | 2020 |
Undrained Anisotropy and Cyclic Resistance of Saturated Silt under Principal Stress Rotation
|
Géotechnique · Q1 | Published |
| CORE-09 | 2015 |
Efficient Robust Geotechnical Design of Drilled Shafts in Clay Using a Spreadsheet
|
J. Geotech. & Geoenv. Eng. · Q1 | Published |
| CORE-10 | 2015 |
Maximum Likelihood Principle and Its Application in Soil Liquefaction Assessment (Ch. 4)
|
Book Chapter | Published |
| CORE-11 | 2014 |
Robust Design in Geotechnical Engineering: An Update
|
Georisk · Q1 | Published |
| CORE-… | … |
Full publication list (28 journal articles, 28 conference papers) available in the CV
|
· | See CV |
* denotes student advisee · Q1 = top-quartile journal (JCR 2024)
PROJECTS
Active investigations.
Six funded awards and three pending proposals, totaling over $827K in funded research since 2019.
Optimizing Field Compaction of Granular Materials
Research and guideline development extending my roller-integrated sensing work toward practical compaction guidelines that state DOTs can adopt for routine use.
Roller-Integrated Acoustic Wave Detection for Continuous Compaction Evaluation
A novel acoustic-wave technique that analyzes vibrational signals from roller operation to infer soil stiffness in real time. With co-PI Dr. Mehdi Norouzi (ECE) and Ph.D. student Kwame Edjah.
Roller-Integrated Image Processing for Compaction Assessment
A parallel investigation exploring computer-vision approaches to subgrade compaction assessment, complementing the FHWA acoustic sensing program.
Machine Learning for Resilient Modulus Prediction of Subgrade Soil
My first external award, which established the agency relationships that led to the two larger sensing projects above.
Investigating Innovations in Litter Collection
Co-PI on an Ohio DOT project led by Dr. Lei Wang.
Connecting Women Faculty in Geotechnical Engineering (GTWF Seed Grant)
My earliest funded work, supporting professional network-building among women faculty in geotechnical engineering.
Sensitivity of Liquefaction Triggering Models to Case-History Composition and the Information Value of Future Reconnaissance
A USGS proposal asking how sensitive empirical liquefaction triggering models are to the composition of the case-history database that calibrates them, and what additional reconnaissance data would be most valuable in improving them. Builds directly on my long-standing work in probabilistic liquefaction assessment and CPT-based hazard characterization.
Applicability of Measurement While Drilling Within Ohio Geological Strata
An ODOT proposal investigating the applicability of Measurement While Drilling (MWD) for characterizing Ohio’s geological strata. MWD delivers real-time, drilling-integrated readings of subsurface conditions, which fits naturally with my broader interest in field-deployable sensing for geotechnical characterization.
Operationalizing the Soil Resilient Modulus Prediction Models: Implementation Phase
A follow-on to my earlier FHWA/INDOT machine learning work that moves the resilient modulus prediction models from research output to deployment. The implementation phase delivers an internal, user-friendly interface that INDOT engineers can use as part of routine pavement design workflows.
& TEAM
The lab.
Research at this depth is not solo work. I currently advise three graduate students and have graduated one Ph.D. and one M.Sc. student, with collaboration that continues well past graduation.
Recently brought to full operational status, our laboratory is anchored by a triaxial testing system equipped with an environmental chamber, capable of controlled temperature cycling and frozen-soil testing. The facility supports a new experimental research thread on the mechanical behavior of soils under thermal loading and freeze-thaw cycles, generating laboratory data that feeds directly into the computational and machine-learning pipelines that drive the rest of the program.
currently advising
currently advising
(1 Ph.D. · 1 M.Sc.)
(3 Ph.D. · 5 M.S.)
TAUGHT
Courses taught.
Sixteen course offerings at the University of Cincinnati since 2019, across the geotechnical and construction-management curricula.
| Course No. | Title | Level | Terms Taught |
|---|---|---|---|
| CVE 3003 | Reliability | Undergraduate | Fall 2022, 2023, 2024, 2025 |
| CVE 5160 / 6060 | Slope Stability | Graduate | Spring 2020 – 2025 (6 offerings) |
| CM 3037 | Soil Mechanics for Constructors | Undergraduate | Spring 2025 (new course, developed) |
| CVE 3002C | Soil Mechanics & Lab | Undergraduate | Fall 2020 – Spring 2022 (4 offerings) |
| CVE 5181 / 6081 | Foundation Engineering | Graduate | Fall 2019 |
| CVE 7092 | MS / PhD Research | Graduate | Ongoing |
Prior teaching at Clarkson University (2017–2019) and Clemson University · full evaluation record in the CV
CONTACT
Service to the profession.
Editorial, reviewing, leadership, and committee work across the profession, the university, and the department.
Leadership & Editorial
- Vice Chair, ASCE Geo-Institute Technical Subcommittee on Risk Assessment & Management
- Secretary, same subcommittee (RAM)
- Editorial Board, Georisk: Assessment and Management of Risk for Engineered Systems and Geohazards
- Awards Secretary, ASCE GI Subcommittee on Earth Retaining Structures
Journal Reviewing
- Canadian Geotechnical Journal · Engineering Geology · JGGE
- Georisk · Transportation Research Record
- Journal of GeoEngineering
Conference Service
- Conference Advisor, Geotechnical Frontiers 2025
- Technical Organizer & Session Chair, Geo-Risk 2023
- Session Chair, GeoCongress 2022
- Bright Spark Lecture, GeoRisk 2023 (awarded by ISSMGE President)
University & Department
- COACHE Survey Committee, University of Cincinnati
- Committee for Department Chair Search, CAECM
- Ad-Hoc Committee on RPT Criteria, CAECM
- Graduate Committee, CAECM
Bright Spark Lecture Award (ISSMGE President, GeoRisk 2023) · Best Paper, ASCE GeoShanghai 2014 · Excellent Paper, Journal of GeoEngineering 2015 · ASCE Middlebrooks and Georisk Best Paper nominations. Recent invited talks include ASCE Continuing Education (2024), the ASCE GI Director’s Cut interview series (2024), Purdue Road School (2024), the TRB AKG20 Committee (2024), and the ISSMGE TC309 international session (2023).
Contact & correspondence.
Office 795 Rhodes Hall, PO Box 210071
2850 Campus Way Drive
University of Cincinnati, Cincinnati, OH 45221
Department Civil & Architectural Engineering and Construction Management
Google Scholar scholar.google.com profile
ORCID 0000-0002-0667-7050
LinkedIn linkedin.com/in/sarakhoshnevisan
Curriculum Vitae Download CV (PDF)