Every home deserves a closer look
| SOCIETAL CHALLENGE | PARTNERS · WHO BENEFITS | AI TECHNOLOGY | EXPECTED IMPACT |
|---|---|---|---|
| Hidden earthquake damage in brick masonry can be missed by visual inspection, leaving families uncertain. | Structural engineers, disaster-response groups and affected communities. Guided by Miyamoto, AIS and Kathmandu University. | Explainable machine learning on vibration from controlled taps, using physics-based features and healthy-wall baselines. | A ~$50 screening unit that helps engineers prioritize areas needing further investigation. |
THE CHALLENGE
Earthquakes destroy homes in seconds but leave families facing months of uncertainty. Even when buildings stand, hidden cracks and weakened brick–mortar connections can go undetected.
Assessments rely heavily on visual inspection, while specialized testing equipment is expensive and needs trained operators. Where engineering resources are limited, that leaves a critical gap. TremOR aims to close it with affordable, objective measurements of masonry condition.
WHY IT MATTERS
PEOPLE
Families who lose homes face more than damaged walls: days of uncertainty without the safety, stability and belonging of home.
IMPACT
Accessible hidden-damage screening gives engineers better information, speeds up investigations and helps communities move toward recovery.
OUR AI SOLUTION
TremOR uses controlled impacts and low-cost sensors to measure how masonry walls vibrate. Materials, moisture and sensor contact also shape these signals, so no single vibration pattern reliably means damage.
Our system combines physics-based signal features with explainable machine learning, built on healthy-wall baselines and tested on specimens it has never seen. It returns interpretable measurements and confidence scores that help engineers decide where to investigate further.
TremOR supports engineering judgment; it does not replace it.
UNDER THE HOOD
Sensors IIS3DWB and LSM6DSV16BX accelerometers · TAL220 load cell with HX711 · DS18B20 temperature probe · solenoid striker on an XY frame · Teensy 4.1 logger
Models nearest-centroid, k-nearest-neighbour and Gaussian classifiers on force-normalised vibration features; whole specimens held out for testing; the model abstains when unsure
HOW IT WORKS
WHERE TREMOR FITS
WHAT WE BUILT
How it works 1 Knock: a striker taps the brick like a small hammer. 2 Listen: sensors feel it vibrate, a “fingerprint”. 3 Compare: software checks it against healthy brick. 4 Point: the engineer sees where to look closer.
PROGRESS SINCE PHASE 1
| PHASE 1 | PHASE 2 PROGRESS |
|---|---|
| Concept and AI pipeline: tap checks, healthy-baseline scoring | Lab bench operating; healthy-brick baseline data collection begun. We aim to identify defect detection accuracy. |
| Interviews with 24+ engineers, researchers and relief professionals | Testing discussions under way with Miyamoto, AIS Construction and Kathmandu University |
PATH TO SCALE
Start with engineer-supervised inspection of earthquake-affected brick. After validation: routine checks of aging masonry, then bridges and industrial sites.
Customers: structural engineering firms, inspection providers, disaster-response organizations and public agencies.
Market context: NDT equipment, $4.7B → $7.4B by 2030. Phase 2 establishes TremOR’s technical performance and practical value.
PHASE 2 GOALS
- Build and verify 2 pilot kits, then 26 more (2 for Kathmandu University)
- 56 test assemblies with known defects, 24 built in Nepal
- 16 reference tests and calibration checks between devices
- Test the AI on unseen specimens against simpler methods
- 3 engineer-led field days: usability, time, agreement with engineers
WORKING WITH PARTNERS
- Miyamoto International: Expert feedback on hidden damage at brick–mortar interfaces; potential field evaluation.
- AIS Construction: Discussed access to a historic brick building for supervised field testing.
- Kathmandu University: Potential collaboration for replicated testing in Nepal. These conversations moved TremOR beyond lower-cost testing toward finding hidden masonry anomalies that might otherwise go unnoticed.
THE TEAM
- Mehdi Badawi (Student Lead)
- Abdullah Mohamed
- Alia Eissa
- Dr. Hazim El-Mounayri (Advisor)
- Mechanical Engineering, Purdue Indianapolis
CONTACT
- badawim@purdue.edu
- mohame49@purdue.edu
- aeissa@purdue.edu