Product
CanGuard
Not a snapshot. The whole movie.
Continuous, millisecond-level recording of the entire CAN bus, with remotely deployed triggers that upload the ribbon around any fault — automatically labelled and ready to train predictive models.
Watch the film
The film could not be loaded.
Traditional software saves a snapshot: the parameters at the exact instant a code appeared. By the time anyone looks, the engine is fine again. A snapshot does not tell what was going on.
The ribbon
CanGuard runs on the Dragonfly and continuously records the entire CAN bus — every parameter, exactly as each ECM publishes it, down to the millisecond.
Triggers
In emd-cloud — also called SIG — a trigger is defined: a fault code, its FMI, its time window. It deploys to one unit — or to the whole fleet. When the fault appears, no one needs to be watching: the event fires the trigger and the ribbon uploads itself, tied to that event.
The black box
If the worst ever happens, the complete ribbon is still on the device. It can be recovered, and the exact post-mortem rebuilt — down to the last millisecond before the damage.
A dataset that builds itself
Every event uploads labelled: which fault, which engine, under what conditions. With no manual work, the whole fleet builds the dataset needed to train models that learn what a failure looks like before it becomes one.