The designer composes the task
Select the target board and measurements. YProbe resolves probe points, expected signals, setup information and conditional branches from the private session.
YProbe packages probe points, expected signals and conditional next steps for a remote technician. The netlist, BOM and diagnostic reasoning stay with the designer; signed results return for local verification and diagnosis.
Remote debugging usually forces a bad choice: send sensitive design context to the factory, or reduce the technician to an unstructured video call. YProbe separates the measurement task from the diagnostic context so both sides receive only what they need.
Select the target board and measurements. YProbe resolves probe points, expected signals, setup information and conditional branches from the private session.
The free technician client shows ordered probe tasks and captures evidence. It does not need the board graph or the AI's ranked hypotheses.
Only allow-listed fields are written. Net names travel as stable per-board aliases and free-text fields are scrubbed before the package leaves the designer.
Results return signed. The designer's YProbe instance recomputes measurement verdicts from the raw samples rather than trusting a remote client's conclusion.
| Included in the task | Structurally absent | Returned to the designer |
|---|---|---|
| Per-board aliased probe point | Original netlist and component graph | Raw captured samples |
| Expected signal and tolerance | Bill of materials | Instrument and capture metadata |
| Instrument/channel setup needed for the task | AI reasoning and ranked hypotheses | Technician notes through scrubbed fields |
| Conditional next-step instructions | Unrelated board and session history | Package signature for verification |
A task queue can include conditional branches: if the first rail fails, take one path; if it passes, move to the next signal. The technician can continue through a prepared measurement tree without waiting for the designer to wake up, while the diagnostic context remains on the designer's machine.
Keep proprietary design context at the design house while giving the factory a repeatable measurement plan for failing units.
Standardise the evidence returned from remote benches so the next diagnostic decision is based on comparable captures rather than photographs and prose.