Trillium Engineering is collaborating with SightX, a deep-tech company specializing in Edge AI, on the integration of SightX’s advanced AI into Trillium’s gimbals to expand the company’s data products offering and further amplify customers’ abilities to execute ISR missions.
SightX’s Merlin software offers an advanced multi-object acquisition engine for mission-critical airborne applications at all wavelengths. Combining these modern processing capabilities with Trillium’s high-end raw video at the edge provides a valuable asset to warfighters by reducing the clutter of information at the ground station and enhancing the ability to deliver critical capabilities in comms-denied environments.
All Trillium gimbals now support the ability to detect, track, and classify objects in both the visible and infrared spectrums with a human ‘on the loop’, instead of having to always be ‘in the loop’.
“This advancement will dramatically enhance the accuracy and efficiency of ISR missions,” said Ryan O’Connor, Trillium’s VP of Engineering. “Our collaboration with SightX delivers improved geolocation target accuracy, greater capacity to derive valuable insights from video imagery, and allows for data processing to occur onboard the UAS. This means operators will only receive important data (rather than all the data) in a much higher-quality format for assessment.”
To date, Trillium gimbals have been equipped with onboard processing to support target locations up to and exceeding Category 1, as well as high-quality video, still image recording, and downlink. In addition, Trillium’s ground station software, SkyLink, provides a powerful user interface to support gimbal control, video and telemetry recording, target tracking, and a 3D map with target breadcrumbs and measurements displayed.
“These features aren’t currently available for Group 2 or 3 UAS gimbals,” said Mark Mirelez, CEO of Trillium Engineering. “Such capabilities have significant military and commercial applications, allowing for POL data, object identification and cueing, and most importantly, allowing for AI processing at the edge in a GPS-denied environment.”
