Engineering Resources • Decision Guide

Remote Monitoring Architecture Guide

Merobix Engineering •

A decision guide for engineers and operations managers designing monitoring for remote assets - wellsites, tanks, pump stations, lift stations, or anything else far from the office. It walks the reference architecture end to end, then works through the decision factors and the cellular vs satellite vs radio trade-offs.

Back to Resources

The Reference Architecture

Nearly every remote monitoring system, regardless of vendor or industry, reduces to the same five stages. Getting the architecture right is mostly a matter of making deliberate choices at each stage instead of inheriting defaults.

Remote monitoring reference architecture Five stages left to right: field devices feed an RTU or edge gateway, which sends data across a communications layer of cellular, satellite, or radio to a monitoring platform, which serves users on web and mobile clients. Field devices Transmitters Sensors PLC / RTU I/O Meters, valves RTU / gateway Protocol translation Local buffering Store and forward Optional local logic Comms layer Cellular Satellite Licensed / unlicensed radio Outbound-only preferred Platform Historian Alarming and callouts Dashboards, reports Cloud or on-premise Users Operations, pumpers Engineers Management Web and mobile Data flows left to right; commands and acknowledgments flow back along the same path where the design allows it.

Decision Factors

Site count and growth

One site justifies almost any ad-hoc solution; fifty sites punish it. The more sites you expect, the more the decision should weight standardized gateway configurations, remote management, and per-site recurring cost over per-site cleverness.

Power at the site

Grid power removes constraints; solar-plus-battery imposes them everywhere. On a solar site the power budget - panel size, battery capacity, and worst-season sun - drives gateway selection, radio choice, polling rate, and even protocol (report-by-exception traffic costs less energy than tight polling). Actual figures are site-specific: build a power budget per site class rather than borrowing numbers.

Comms availability

Survey before you standardize. Cellular coverage maps are optimistic at antenna height on a tank battery in a draw; radio needs line of sight or engineered paths; satellite works nearly anywhere with sky view but changes the cost and latency profile. Mixed fleets commonly end up with a primary technology plus exceptions.

Latency and update needs

Be honest about what the operation needs rather than what sounds impressive. Daily production accounting, hourly tank levels, and prompt alarm callouts have different requirements, and the strictest one sets the design. Report-by-exception protocols can deliver prompt event notification even on modest links.

Security constraints

Common practice for cloud-connected sites is outbound-only connections initiated by the gateway, TLS on every hop that supports it, and no inbound ports open to the field network. Corporate policy may additionally require private APNs, VPNs, or on-premise hosting. The zones-and-conduits concepts of the ISA/IEC 62443 series and the guidance in NIST SP 800-82 are the standard reference points for structuring this.

Budget structure

The technologies split less on total cost than on cost shape: radio is capex-heavy (you build and own the infrastructure), cellular and satellite are opex-heavy (recurring service per site). Which shape fits depends on your ownership horizon and how your organization funds projects.

Cellular vs Satellite vs Radio

FactorCellularSatellitePrivate radio (licensed or unlicensed)
CoverageWherever carriers have built; verify at each site, not on the mapNearly anywhere with a clear sky viewWherever you engineer it: line of sight or designed paths to your own infrastructure
Infrastructure you ownGateway and antenna onlyTerminal and antenna onlyRadios at every site plus repeaters or towers - yours to build and maintain
Cost shapeRecurring service per siteRecurring service per site, historically at a premium for airtimeCapex up front; low recurring cost, plus licensing fees where a licensed channel is used
BandwidthGenerally ample for SCADA telemetryVaries widely by service class; plan around the plan you buyModest by modern standards; well matched to report-by-exception telemetry
Latency characterLow for telemetry purposesDepends on constellation: geostationary links add noticeable round-trip delay from orbital distance; low-earth-orbit services are lowerLow on a healthy path
Typical failure modesTower outages, congestion, carrier sunsets of older network generationsWeather fade, antenna obstruction or misalignmentPath obstruction (growth, new structures), interference on unlicensed bands, your own tower maintenance
Best fitFleets of dispersed sites inside coverage; fastest to deploySites outside terrestrial coverage; backup path for critical sitesDense clusters of sites you will operate for years, especially where coverage is poor

Patterns That Tend to Work

Sources & Standards

Primary references for securing and structuring remote monitoring architectures:

From Diagram to Dashboards

This guide is complete without us - use it to pressure-test any vendor, including this one. If the platform stage is the part you would rather not build, that is what Merobix is.