How many devices should I start with before scaling? — answered by Engineer Saeed Baatiah
One working completely, then three to five in the real location rather than on a bench. The second stage is what reveals what a lab never shows: network coverage in the building’s corners, heat and dust and their effects, how the team behaves with alerts, and what installation actually involves and how long it takes. After two weeks of real operation you know the true per-unit cost, and only then does ordering a hundred become a decision rather than a gamble.
A pilot in the real location reveals what a bench never shows: network coverage in the building’s corners and behind metal walls, the effect of heat, dust and humidity on both readings and hardware, the actual installation time per unit (usually several times the estimate), and how your team behaves with alerts — whether they respond or have learned to ignore them within a week.
After two weeks of real operation you have numbers rather than estimates: true per-unit cost including installation, the share of units that needed intervention, actual battery consumption, and the daily alert count. Those four turn ordering a hundred devices from a gamble into a calculation — and may reveal that the right number is twenty rather than a hundred, which is itself a large saving.
- One on the bench, then three to five in the real location
- Real installation time is usually several times the estimate
- After two weeks you know the true per-unit cost