Electrical hazard analysis, arc early warning, device health diagnosis, and data generation from hundreds of types of smart hardware.
Industrial AI Analysis System
Industrial AI Analysis System | Knowledge Centre
Electrical hazard analysis, arc early warning, device health diagnosis, and data generation from hundreds of types of smart hardware.
Related Knowledge
Want a deeper look at FEXLINK solutions?
Contact the FEXLINK solutions team for customised solutions and technical support.
Frequently Asked Questions
Which functions make up the Industrial AI Analysis System category?
Three named functions: electrical hazard analysis, arc early warning and device health diagnosis, all fed by data generated from hundreds of types of smart hardware. The analysis is only as specific as the hardware fleet behind it, which is why data generation leads.
Which 3 measurement topics feed the diagnostic analysis?
Electrical signal capture, virtual-power-plant data and SPD online monitoring. Each supplies a different input the diagnosis consumes - waveform detail, dispatch context and protection state - and the weakest of the three caps how confident any conclusion can be.
How does a diagnostic engine tell an arc from a normal load disturbance?
By combining low-frequency and high-frequency features, half-cycle-level waveforms and scenario filtering rather than one threshold. With capture reaching a 100MHz range, switching noise and drive interference get filtered while recurring arc pulses accumulate evidence across time scales.
Why does device health diagnosis need correlation instead of single limits?
A single limit reports an effect - temperature high, current high - never the cause behind it. Correlating temperature, current and leakage over consecutive periods identifies growing contact resistance, so the output is a diagnosis with a repair action, not a number.
What happens when arc early warning is built on one high-frequency band?
False alarms pile up, because switch operations, motor starts, adapters and variable-frequency drives produce look-alike pulses. The fix is multi-band, multi-time-scale evidence with scenario filtering, which is why the diagnosis layer sits above the raw 100MHz capture.