SWIR and thermal imaging sensors technology and investment research

Technical bottleneck

• Bottleneck type: Manufacturing yield / Export control
• Technical constraint: InGaAs epitaxial layer uniformity across large-format focal plane arrays (1280×1024+); SWIR dark current at room temperature degrades SNR — cooling improves performance but adds size/weight/power; microbolometer pixel pitch approaching 8 μm faces thermal crosstalk limits
• Economic constraint: InP substrate supply chain (see Layer 1); ITAR/EAR restrictions on high-resolution SWIR FPAs; USG investment in domestic InGaAs foundry capacity

Adoption

• Driver: Autonomous vehicle perception fusion requiring all-weather/all-light sensing; defense SWIR target identification and ISR; industrial SWIR for semiconductor inspection and agricultural sorting
• Blocker: LiDAR and high-resolution visible cameras handling adverse weather; cooled MWIR sensors (more expensive but higher sensitivity); Chinese InGaAs sensor production

Public companies exposed

FLIR (acquired by TDY)

RTX (Raytheon Vision Systems)

LMT (Santa Barbara Focalplane)

TSEM (Tower Semiconductor — InGaAs foundry)

Validation signals

SWIR design wins on autonomous vehicle platforms; defense SWIR FPA production contracts; uncooled SWIR sensitivity improvements reducing size/cost

Invalidation signals

Cooled MWIR replacing SWIR for long-range applications; thermal camera commoditization (Chinese microbolometers); event-based cameras capturing adverse-weather perception