ORIGIN
The case for better
Edge AI systems
Today's unmanned systems depend on vulnerable infrastructure and constant human supervision.
To be useful members of hybrid teams, they must operate reliably at the edge, where the risk is highest.
Many of our team have worked in these environments, and that experience shapes what we build.
This understanding has allowed us to frame difficult questions, such as:
01
How do we build remote sensors that alert only to genuine threats, with a low false-positive rate to stay credible over months of deployment?
02
How do we alert on military vehicles of interest while ignoring normal traffic, without missing the vehicles that matter, ensuring a low false-negative rate?
03
How does a small UAS navigate accurately through a multi-hour mission, without reference to any external system or infrastructure?
04
How does a small UAS recognise a change on the ground, find the critical information and report it immediately, rather than continuing on its route?
IMPACT
Our edge AI systems answer these questions.
They keep people out of harm's way and take over the work of watching sensors, so people can focus on decisions.
