A pilot is useful when it answers a decision. Before installing devices, define the service requirements, representative locations and evidence needed to decide whether to expand, change the design or investigate further.
Define the population and window
A read-rate percentage is incomplete without the number of eligible meters and the collection window. Decide whether the target is a daily report, a billing read or an event delivery. Record exclusions and avoid changing the population after seeing results. Compare the same definitions across alternative technologies.
For a WE RICH product comparison, the LXC DN15–DN50 ultrasonic water meter with valve control combines ultrasonic measurement with remote valve control for residential supply points.
For a WE RICH product comparison, the LoRaWAN gateway forms part of the communication path between compatible meters and the application; coverage and backhaul must be planned for the site.
For a WE RICH product comparison, the Head End System (HES) brings collected readings and communication status into an operational view, with interfaces agreed for the project.
Measure more than reception
Track valid account-linked readings, metrological test results, installation defects, latency and field interventions. A message can arrive with the wrong unit or account mapping. A local register can work while remote delivery fails. Separate these outcomes so the project can fix the correct part of the system.
Agree the decision rule
Document thresholds, test duration, responsibilities and the response to failure before the pilot begins. Include recovery after outages and representative difficult installations. Label simulated scenarios clearly. Report observed results with limitations and unresolved exceptions. A pilot report should explain what was demonstrated, not project unmeasured savings or universal performance.
Practical checklist
- Freeze definitions before testing.
- Include difficult installations.
- Report unresolved exceptions.
Frequently asked questions
Is 100% reception a complete KPI?
No. Define the population, interval, validation and business use.
Should failures be excluded?
Only use exclusions agreed in advance and disclose them.
Does a pilot guarantee full-scale results?
No. Assess how scale and location differences affect the conclusion.
Related articles
References and technical reading
Product capabilities depend on the model and its documentation. Confirm installation and performance requirements for the project.