Target Node AI Anomaly Detection
Your robot knows when something is wrong.
Now you will too.
Target Node Anomaly Detection learns what your robot's sensors look like when everything is working. When the pattern breaks, it alerts you and tells you what's wrong.
Watch the demo
The problem with thresholds
A common way to monitor robot health is to set min/max limits on each sensor and raise an alarm when a reading goes out of range. This works for big, obvious failures. However, it misses a whole class of faults where every sensor can read as normal individually but the relationships between the readings are anomalous. These faults are very difficult (or impossible) to catch with manual thresholds.
Target Node AD learns what the sensors normally look like together, so it detects faults even when all individual readings appear normal. All without setting or tuning thresholds.
How it works
See how it works through real examples
Examples with injected faults on simulated robots. Every sensor reads normal on its own. The fault is detected anyway.
Good to know
Runs on the robot
Detection happens on the edge. No streaming your data to the cloud.
Protocol Agnostic Architecture
Works with ROS 1 and 2 standard and custom message types out ofthe box. Can be extended easily to ADS, OPC UA, CAN and others.
Lightweight
A few percent of one CPU core and tens of megabytes of memory.
Shows you where
Alerts include the specific signals that changed, so you know what went wrong.