Pilot Program
Test the monitoring question before scaling the system.
A useful pilot makes the asset, operating states, evidence boundary and decision path explicit. It should be able to support a proceed, refine or stop recommendation.
Pilot principle
One asset question. Defined evidence. An accountable decision.
Stage-gated process
Five phases, with a decision at every boundary.
The exact technical scope is confirmed only after the application review. The sequence below sets expectations without promising a predetermined result.
- 01
Application review
Output: Initial fit and evidence questions
Identify the asset, operating envelope, access constraints and maintenance decision that the pilot should inform.
- 02
Site and data definition
Output: Pilot boundary and data plan
Map candidate sensing locations, representative operating states and the existing data needed for correlation.
- 03
Acceptance design
Output: Acceptance criteria and review path
Agree how observations will be reviewed, what constitutes useful evidence and which claims are explicitly out of scope.
- 04
Pilot execution
Output: Traceable pilot evidence
Install the agreed inputs, collect representative observations and document operating context throughout the evaluation.
- 05
Scale recommendation
Output: Proceed, refine or stop recommendation
Review fit, limitations, integration needs and the evidence required before any broader deployment decision.
Good pilot inputs
Bring the operating reality into the first conversation.
- A named asset or route in a defined operating context
- A maintenance decision that current evidence does not fully resolve
- Access to operators, maintainers and relevant existing records
- Agreement on safety, site access and data-governance boundaries
Not a fit without further definition
A pilot cannot validate an undefined promise.
- Requests for a universal accuracy guarantee before site review
- Safety-critical replacement of certified protection systems
- No access to representative operating states or validation evidence
- A requirement for autonomous root-cause diagnosis without engineering review
Before we start
What we will ask.
- Asset and consequence
- Which component or route is in scope, and what decision makes it important?
- Operating states
- Which loads, speeds, modes or environmental conditions define normal operation?
- Existing evidence
- What inspection, vibration, process or maintenance records can support correlation?
- Decision and acceptance
- Who reviews the evidence, and what would make it useful enough to continue?
Application intake
Have an asset and decision in mind? Start the review.
Share the asset, operating context and maintenance decision you want to improve. We will help frame a practical pilot path.