Sensors

last updated 2026-05-04 · +14 sources in last 30d

Physics / mechanism

Sensors transduce a physical quantity (pressure, photons, temperature, chemical concentration, acceleration) into an electrical signal. The core physics depends on the modality: piezoresistive/capacitive for MEMS pressure/inertial, photoelectric effect for image sensors and LiDAR receivers, electrochemical for gas/biosensors, pyroelectric/bolometric for thermal IR. Key parameters: sensitivity (signal/input unit), noise floor (NEP for optical, μg/√Hz for IMUs), bandwidth, drift, and power. State of the art: Sony/Samsung BSI CMOS image sensors at sub-micron pixels; Bosch MEMS IMUs at <10 μg/√Hz noise; uncooled VOx bolometers at NETD <30 mK. Integration trend is sensor fusion on-chip with edge inference.

Competitive landscape

MEMS sensors (Bosch, STMicro, TDK/InvenSense) dominate inertial/pressure at commodity margins. Photonic sensors—silicon photonics-based LiDAR, integrated spectrometers, optical gyros—compete on performance-per-power at higher ASP. Emerging: quantum sensors (atomic interferometry, NV-center magnetometers) threaten MEMS in navigation accuracy but remain pre-commercial. Chemical/biosensors fragment by vertical. The meaningful battleground is CMOS-integration density vs. heterogeneous packaging of III-V or LN-on-Si for photonic modalities.

ModalityKey metricIntegration path
MEMS inertialNoise floor μg/√HzMonolithic CMOS
Photonic (LiDAR/spectro)NEP, angular resSi photonics / III-V flip-chip
Quantumppb sensitivity, driftHybrid, nascent fab

Companies using

Connected ideas

Sources

Frontier (open questions)

Recent mentions

Frontier questions