Embedded Systems

Build a passive radar system that uses FM or DVB-T signals to detect moving objects, then analyze cross-ambiguity data and signal processing performance.

Model a smart irrigation network, test packet loss, and measure how co-simulation changes reliability, latency, and control performance.

Build a vibration sensor system that classifies motor faults with FFT and CNNs, then compare public bearing data with your own measurements.

Build an adaptive Li-Fi link, measure ambient-light noise, and compare OOK, PAM-4, and OFDM for higher throughput in real rooms.

Build a speckle-based biometric system and test TinyML classification, signal quality, and identity accuracy with real sensor data.

Build a receiver pipeline that flags spoofed ship and aircraft positions by comparing broadcast data with multilateration checks.

Build and test a USB dongle that changes keystroke timing patterns, then measure how well it disrupts biometric profiling with statistics.

Build a privacy-preserving sensor system that estimates engagement from heat patterns, then test accuracy, latency, and on-device inference.

Design a piezoelectric tile and test how rectifier choice, waveform classification, and supercapacitor buffering change harvested energy and output stability.

Build a frequency-sweep impedance classifier with an ESP32 and learn calibration, curve fitting, and ML for counterfeit capacitor detection.