"How long does this actually take to set up?" comes up in nearly every first conversation with a facility manager — usually followed by some skepticism, because most enterprise software rollouts stretch into weeks. IoT hardware installation is a different kind of project: mostly physical work with a well-defined sequence, not a software integration project with unknown unknowns. Worth walking through what actually happens, hour by hour and protocol by protocol, instead of just quoting a number.
Before the Team Arrives
A short site survey call covers what's needed before anyone shows up: equipment specs and manufacturer details, which interface the controller exposes (Modbus RTU/TCP, 4-20mA analog, or RS-485 — most controllers built in the last two decades support at least one), panel access points, and a rough sense of cellular signal strength on-site. This is also when mounting locations get provisionally agreed on, so the on-site visit is execution, not exploration.
Hours 0-4: Site Survey and Sensor Placement
The on-site team confirms exact mounting points against what was discussed on the survey call, checks cellular signal strength at each specific location (not just the site generally — a generator yard's signal can differ meaningfully from the office fifty meters away), and physically installs sensors — current clamps on the alternator or transformer leads, temperature probes on the engine block or oil line, pressure transducers on the boiler's steam line, whatever the specific asset type requires. For a single-site, single-asset installation, this is typically the bulk of the on-site time: 2-4 hours, and equipment stays running throughout.
Hours 4-24: Wiring, Mounting, and Connectivity Verification
Sensors get wired into the gateway over the confirmed protocol — Modbus RTU/TCP, 4-20mA, or RS-485 — and mounted permanently. Each device is then verified against the cellular network to confirm it's transmitting, encrypted with 256-bit AES in transit. This stage often runs partly unattended: devices left reporting overnight is a natural checkpoint for confirming connectivity is stable — including a deliberate test that the local buffer correctly queues readings if the signal drops briefly — before moving to dashboard configuration.
Hours 24-40: Dashboard Configuration and Threshold Calibration
With live data flowing, the dashboard gets configured for your specific site — asset naming, user roles (superadmin, admin, or site-scoped monitor access, depending on who needs to see what), and initial alert thresholds based on the equipment's rated specs from its nameplate. These starting thresholds are a baseline, not a final answer; they get refined over the following one to two weeks as your equipment's actual operating pattern — not just its rated capacity — becomes visible in the data.
Hours 40-48: Test Alerts, Handover, and Training
Before handover, the team triggers test alerts end-to-end — confirming WhatsApp/SMS notifications actually reach the right people's phones, not just that the dashboard shows a number internally. A short walkthrough covers how to read each of the eight tracked parameters for your asset type, what the initial thresholds mean, and who to contact during the calibration period that follows.
What Happens If Something Doesn't Go to Plan
Real sites vary — a panel that's harder to access than described, a cellular dead zone in one specific corner that needs an external antenna, equipment old enough that it doesn't expose a standard Modbus or RS-485 interface and needs a different sensor approach entirely. When that happens, the 48-hour estimate stretches, but the sequence doesn't change: survey, install, verify, configure, hand over. The goal isn't a rigid deadline — it's a predictable, protocol-agnostic process from first site visit to a dashboard you can trust, regardless of how standard or unusual your specific equipment turns out to be.
Frequently Asked Questions
For most standard installations, sensors are mounted alongside running equipment without requiring it to be shut down — current clamps and temperature probes are non-invasive, and the 4-20mA/RS-485 tie-in points on most modern controllers are accessible without an outage. The exact approach depends on the specific equipment and mounting points, which the site survey call covers before the visit.
Monitoring is cellular-first (3G/4G/5G) by default specifically for this reason — the installation doesn't depend on your facility's Wi-Fi being available or reliable, and the gateway's 30-day local buffer means even a poor-signal location doesn't lose data, just delays how quickly it reaches the dashboard.
Typically whoever has equipment access and authority to approve sensor placement — a facility engineer or maintenance lead is usually enough; it doesn't require your full team on-site.
The dashboard is live and alerting on baseline thresholds set from the equipment's rated specs — from there it's a calibration period, typically one to two weeks, where thresholds get tightened based on your equipment's actual operating pattern rather than the generic starting point.
Modbus RTU/TCP, 4-20mA analog, and RS-485 digital — the standard interfaces most industrial equipment controllers already expose, regardless of manufacturer. The gateway aggregates all of them and transmits over cellular or Ethernet with 256-bit AES encryption, so nothing about the wiring is proprietary or locks you into a specific brand of equipment going forward.