High-speed optical inspection of a circuit board, similar to the technology used in the Cheetah HS sensor. Photo: AIStorm
GIF Promo

California-based AIStorm and Israel’s Tower Semiconductor are pushing battlefield imaging into overdrive with the Cheetah HS sensor, embedding AI directly into the chip to capture up to 260,000 frames per second.

The 120×80-pixel device integrates AI at the first processing layer, enabling real-time analysis while significantly reducing energy consumption and material costs.

The system leverages Tower’s charge-domain imaging platform and AIStorm’s analog neurons to convert photons directly into electric charge and process data analogically, bypassing traditional high-speed data converters.

According to the companies, this design delivers up to 4,000 times faster than conventional complementary metal-oxide-semiconductor (CMOS) sensors, a potential edge in missions that demand rapid data capture.

Front and back views of the Cheetah HS high-speed AI imaging sensor. Photo: AIStorm

Other features include adjustable frame rates, built-in LED driver, improved low-light performance, and outputs designed for advanced neural processing.

These capabilities can enhance surveillance, targeting, and reconnaissance in environments where every millisecond counts.

“This solution makes slow-motion, real-time analysis accessible to a wide range of markets,” AIStorm CEO David Schie said. “Tower is a global leader in charge-domain pixels with a global shutter, making it the ideal partner for developing breakthrough products.”

Beyond Defense

The chip can deliver either raw high-speed video or pulse streams tailored for neural networks, giving developers flexibility in system design. This versatility extends beyond defense, opening the door to robotics, structural monitoring, and manufacturing.

It can also be used for barcode scanning, printed circuit board inspection, biometric access systems, vehicle-speed detection, and sports analytics.

Cheetah HS is available in both chip form and full reference-camera systems.

You May Also Like

‘Lidar for Cars’ Inventor Unveils Tech to Track Mach 1 Drones

David Hall’s Cam-Lidar merges lidar, AI, and advanced optics to track Mach 1 aerial threats with real-time 3D precision and reliability.

CH-47F Chinook Achieves First Pilot-Free Landing With Boeing Control System

Boeing’s DAFCS software enabled a CH-47F Chinook to complete a fully automated landing with zero pilot input, marking a milestone in autonomous helicopter flight control.