Laboratory
Experiments
Real-world signal capture, resonance analysis, and boundary information studies. Each experiment generates data that feeds into our ML pipelines and reconstruction algorithms.
1
Active
1
Completed
1
Planned
3 experiments registered
Arduino Signal Capture #001
Four-channel vibration signal acquisition from a steel beam structure using piezoelectric accelerometers connected to an Arduino Mega 2560 with external 16-bit ADC. Sampling at 8 kHz across all channels simultaneously. Objective: capture spectral signatures under controlled impulse excitation to build a baseline dataset for structural identification.
Motor Resonance Analysis
Comprehensive spectral characterization of a brushless DC motor under varying load conditions (0%, 25%, 50%, 75%, 100% rated load). Tracking resonance frequency shifts and harmonic content changes as indicators of structural wear and bearing degradation. Dataset contains 4,891 spectral snapshots collected over a 72-hour continuous monitoring session.
Boundary Information Experiment
Validating the boundary-to-volume information encoding hypothesis using 3D-printed test structures with known internal geometry. Acoustic sensors placed at boundary positions will capture impulse responses. The captured boundary data will be fed to a spectral encoder to attempt structural reconstruction of the interior.
Experiments generate signals. Signals become datasets. Datasets train models. Models reconstruct structure.