
Lantronix and Latent AI have partnered to integrate automatic target recognition and persistent tracking capabilities into Lantronix’s Spotter robotics platform, targeting Group 1 (UAS). Latent AI’s Captus Perception will add target detection and persistent tracking, and the companies are working toward a packaged solution for Group 1 UAS designed to deliver onboard perception within the size, weight, power, and compute constraints of small aircraft. Captus Perception will be available as a Spotter package in the first quarter of 2027, following Spotter’s availability in the fourth quarter of 2026.
On the battlefield, detecting a target is only the first step. Small drones must hold a track as targets move or disappear behind obstacles, often with limited compute and unreliable links. Simple pixel-lock tracking on low-end boards can easily lose the target. More capable perception models exceed the compute, power and cost budgets of small, often attritable aircraft. Spotter was built to close that gap, enabling the combination of its compact, power-efficient compute with Latent AI’s hardware-optimized perception software to support persistent onboard tracking.
With Captus Perception on Spotter, an operator can designate a target and have the drone track it onboard. If the target moves out of view or is obscured, the drone can reacquire it. Tracking data will feed directly to the flight stack to support reconnaissance, surveillance and, where the platform supports it, terminal guidance. The track is designed to hold even when bandwidth is limited or the link is lost.
“Spotter gives drone makers the right amount of AI at the right size, power and cost. Latent AI shows what that unlocks on a Group 1 aircraft,” said Saleel Awsare, president and CEO of Lantronix. Holding a target when it moves, hides or the video link drops is exactly what small drones have been missing. Together, we give customers a mission-ready path to that capability, and we show that Spotter is the platform mission software is built on.”
Why Spotter
Robust tracking on a small drone takes efficient AI software and compute sized for the aircraft. Powered by the Qualcomm® Dragonwing™ QCS6490, Spotter delivers up to 12.5 TOPS of AI performance at single-digit watts on a 42 x 120 mm board. It enables the combination of AI, vision, flight-grade sensing and vehicle control on one platform. It supports PX4, ArduPilot, Betaflight and MAVLink, so tracking data can flow straight to the flight stack. Spotter is NDAA and TAA compliant, is designed for a 10-plus-year lifecycle and builds on Lantronix compute already flying in Blue UAS-approved drones. Latent AI optimizes its AI pipeline for the target hardware, mapping each stage to the right compute engine on the chip. That lets more capable perception models run within Spotter’s power and processing budget.
“Drone makers shouldn’t need a perception team to put real target tracking on a small aircraft,” said Jags Kandasamy, CEO of Latent AI. Together, Spotter and Captus Perception will give them compute and tracking in one package: detection and tracking that runs onboard a Group 1 aircraft and locks onto whatever target the operator picks.”



