The praying mantis: the original edge processor. 50 ms strike speed, 360° awareness, zero-latency detection — reimagined as a family of cameras that understand what they see.
Cheap, commoditised drones have turned low-altitude airspace into a blind spot. Mantis-Edge closes it with passive, on-device AI sensing — cameras that detect, classify and track at the edge, with no cloud, no emissions, and no latency.
An Edge-AI passive unmanned ground sensor engineered for rapid, reliable detection and classification of rotary drones. Inspired by the mantis' high-throughput visual system, the X-01 pairs a high-resolution optical array with dual neural accelerators and multi-modal sensing — producing precise, timestamped track data at the sensor edge.
192 MP stereo-triplet optical array with 10× digital zoom, image stabilisation and IR capability — a visual system built for wide-area aerial awareness.
Dual neural accelerators run multiple AI inference instances simultaneously — 55 trillion operations per second, with a dual-algorithm strategy that holds situational awareness before a threat ever appears.
AI-inferenced audio and EM detection with threat classification, fused with optical tracks into secure, standardised telemetry — position, velocity trends and classification confidence, ready for authorised command systems.
Built for rugged field use, low power and secure operation — for infrastructure operators, public safety agencies and defence customers.
From custom hardware prototyping to specialised AI deployment, we help you transition from simple data collection to real-time, actionable intelligence.
High-performance camera hardware tailored to mission requirements, prioritising optimised SWaP-C for drones, robotics and remote installations.
Built-in AI accelerators for real-time inference with no cloud dependency — low-latency response, autonomous reliability in contested zones, and enhanced privacy.
Beyond motion detection: algorithms trained to distinguish complex subjects from environmental noise, drastically reducing false alarms by interpreting the context behind the movement.
100% Australian-owned and operated. Seamless integration with existing infrastructure, with locally developed and supported intellectual property strengthening Australia's sovereign industrial base.
A complete, plug-and-play platform giving students hands-on experience with real-time edge-AI vision — from high schools and technical colleges to universities and postgraduate research. Full Linux OS, direct code access, and three preloaded inference models: pose estimation, instance segmentation and depth perception.
Mantis-Edge AI is a 100% Australian-owned and operated venture, committed to strengthening Australia's sovereign industrial base with locally developed and supported intellectual property.
"Cameras that understand what they see, so they can change your world."
A strong research association with the Advanced Drone Systems Research Centre — advancing edge-AI aerial sensing together.
Mantis-Edge (in association with the Whistling Dog Drone project) presented at the NSW Commercialisation Showcase, held at the University of Technology, Sydney — May 14, 2026. Watch the presentation.
