.png)
Remote Flow Monitoring with 4G, LoRaWAN and RS485 Architecture
- 52 minutes ago
- 9 min read
A flow meter in a remote pit, pump shed, irrigation channel or mine site is only useful if its data reaches the people who need it. The meter may be accurate, but if readings stay on a local display, teams still rely on site visits, manual logs and delayed fault detection.
Remote flow monitoring solves that problem by connecting field instruments to a cloud dashboard and alarm system. A practical architecture uses RS485 and Modbus at the instrument layer, a Dragino or RTU device at the edge, then LoRaWAN or 4G for backhaul into ProSight.
The full path looks like this:
Flow Meter → RS485/Modbus → Dragino/RTU → LoRaWAN or 4G → ProSight → Dashboard & Alarms
This setup suits water networks, irrigation, mining, pump stations, remote pipelines and wastewater sites where flow rate, total volume and fault status need to be visible without sending someone to site.
The complete remote flow monitoring architecture
A good monitoring system starts with a simple idea: each layer should do one job well.
The flow meter measures water or liquid movement. RS485 and Modbus carry that measurement out of the instrument. The Dragino or RTU device reads the Modbus registers and packages the data. LoRaWAN or 4G sends it to the cloud. ProSight displays the readings, records trends and triggers alarms.
That separation makes the system easier to build, test and maintain.
Layer | What it does | Common field choices |
Flow meter | Measures flow rate, total volume and status | Electromagnetic, ultrasonic, mechanical or insertion meters |
RS485/Modbus | Sends digital readings over cable | Modbus RTU over RS485 |
Dragino/RTU | Reads registers and prepares telemetry | Dragino RS485 node, industrial RTU, data logger |
LoRaWAN or 4G | Sends readings from site to platform | Private LoRaWAN, public LoRaWAN, 4G LTE |
ProSight | Stores, displays and alarms on data | Dashboards, trends, thresholds, notifications |
The key is to avoid treating the communications device as an afterthought. Power, enclosure rating, antenna placement, polling rate and fault handling matter as much as the meter itself.
Start at the flow meter
Most industrial flow meters provide more than one output. A basic pulse output can report volume, such as one pulse per litre or kilolitre. A 4 to 20 mA signal can report live flow rate. For richer telemetry, RS485 with Modbus RTU is usually the better option.
With Modbus, the monitoring device can read values such as:
Instantaneous flow rate
Forward total
Reverse total
Net total
Meter status
Empty pipe status
Error codes
Battery or supply status, if supported
The exact register map depends on the meter. Before wiring anything, collect the meter manual and confirm the following:
Modbus slave address
Baud rate
Parity and stop bits
Register addresses
Data type and byte order
Units and scaling
A common commissioning issue is not a faulty device, but a mismatch in Modbus settings. If the meter uses 9600 baud, even parity and slave ID 1, the RTU must use the same settings. If the platform expects litres per second but the meter reports cubic metres per hour, the dashboard needs scaling.
Good field practice is to test the meter locally with a Modbus tool before connecting the telemetry device. That confirms the instrument responds and the register map is correct.
Use RS485 and Modbus for dependable field wiring
RS485 is common in industrial sites because it handles distance and electrical noise better than many simple signal connections. It uses a differential pair, often labelled A and B, D+ and D-, or positive and negative depending on the manufacturer.
For remote flow monitoring, RS485 offers several useful advantages:
It can carry multiple measurements through one cable
It supports longer cable runs than many low-voltage signal types
It can connect to a wide range of industrial meters and controllers
It gives digital values instead of analogue estimates
Cable quality still matters. Use twisted pair cable, keep it away from high-voltage runs where possible, and follow the manufacturer’s recommendations for shielding and earthing. On longer runs, termination resistors may be needed at the ends of the RS485 bus. Biasing may also be required in some networks to hold the line stable when no device is transmitting.
For a single flow meter connected to one Dragino or RTU device, the wiring is usually simple:
Connect RS485 A to A, and B to B
Connect signal ground if the device manual recommends it
Set the Modbus address and serial parameters
Read the target registers from the RTU
Confirm live values against the meter display
Some labels are confusing because different vendors name A and B in opposite ways. If the device does not respond, swapping the A and B wires is a quick and safe test, assuming the rest of the wiring is correct.
Add a Dragino or RTU device at the edge
The edge device sits between the wired instrument and the wide area network. In this architecture, that device may be a Dragino RS485 product, an industrial RTU or a small data logger with Modbus support.
Its job is to poll the flow meter and send selected values to ProSight. It may also buffer readings during a network dropout, manage power modes, and report its own health.
The device should support the needs of the site, not just the data sheet. In the field, check:
Input support
It must read Modbus RTU over RS485. If pulse or analogue inputs are also needed, the hardware should support them.
Power supply
Some sites have mains power. Others rely on solar and battery systems. Sleep modes and low current draw matter for remote installations.
Enclosure and environment
Outdoor sites need suitable IP-rated enclosures, cable glands, surge protection and safe mounting.
Antenna options
A poor antenna position can make a good device look unreliable. A remote antenna may be needed if the enclosure is metal or installed in a pit.
Configuration access
The team needs a practical way to set polling intervals, register maps and network credentials.
A typical configuration may poll the flow meter every minute, then send a summary every five or fifteen minutes. Critical alarms, such as no flow during pump operation or very high flow, may need faster reporting.
There is always a trade-off. More frequent polling and transmission gives faster visibility, but uses more power and network capacity. Battery-powered sites need a careful balance.
Choose LoRaWAN or 4G for the backhaul
Once the RTU has the data, it needs a path to ProSight. LoRaWAN and 4G both work, but they suit different site conditions.
LoRaWAN is useful when many low-power devices need to send small amounts of data over long distances. It can work well for flow, level, pressure and status readings. It is often a strong fit for farms, irrigation districts, council assets, dispersed water infrastructure and private networks.
4G is useful when sites have mobile coverage and need direct IP connectivity, higher data rates or simpler deployment without a nearby LoRaWAN gateway. It suits pump stations, remote pipelines, mining assets, water treatment sites and temporary monitoring.
Choice | Best fit | Watch points |
LoRaWAN | Low-power telemetry across spread-out assets | Gateway coverage, payload size, duty cycle, antenna height |
4G | Sites with mobile coverage and more frequent data needs | SIM management, signal strength, data plan, power draw |
Hybrid | Mixed networks across a large region | Device standards, data consistency, support process |
For LoRaWAN, the device sends packets to a gateway, which forwards them to a network server and then into ProSight. For 4G, the RTU or gateway usually sends data through the mobile network using MQTT, HTTPS or another supported protocol.
In both cases, plan the communications path before installation. Field signal checks are better than assumptions made from a coverage map. A working 4G phone at site does not always mean a modem inside a metal cabinet will perform well. A LoRaWAN gateway on a low building may not cover a meter inside a below-ground chamber.
Send the data into ProSight
ProSight becomes the operational layer. Field data becomes useful when operators can see trends, compare sites, and receive alarms before problems grow.
A well-designed dashboard should show the current state first:
Live flow rate
Daily, weekly and monthly total
Device online or offline status
Last update time
Alarm state
Battery or supply voltage
Signal quality, where available
Trends matter just as much as live values. A sudden flow spike may point to a burst pipe. A slow fall in flow may suggest a blocked filter, a failing pump or a partially closed valve. Night flow in a water network may reveal leaks. Repeated zero-flow alarms may show pump control problems.
ProSight can also help separate field faults from process faults. If the meter value is flat but the RTU is online, the issue may be process related or meter related. If the whole site stops reporting, the issue may be power, antenna, gateway, SIM or network coverage.
The best dashboards avoid clutter. Operators should not need to search through dozens of raw values to answer simple questions:
Is the site flowing?
Is the flow within the expected range?
How much volume moved today?
Did the device report recently?
Is there an alarm that needs action?
Build alarms that match real operations
Alarms should be useful, not noisy. If every minor variation sends a notification, people stop trusting the system.
Start with a few high-value alarm rules:
High flow
Detect possible bursts, stuck valves or uncontrolled discharge.
Low flow
Detect pump failure, blocked lines or supply interruption.
No flow when expected
Compare flow against a schedule, pump run signal or operating window.
Flow when not expected
Detect leaks, unauthorised use or valve bypass.
No data
Alert when the device has not reported within an expected time.
Low battery or low supply voltage
Plan maintenance before telemetry fails.
Alarm delays are useful. For example, a low-flow condition may need to persist for several minutes before it becomes an alarm. That prevents nuisance alerts from short pump transitions or valve changes.
Use clear alarm names. “Pump Station 4 No Flow During Run” is better than “Alarm 17”. Include the site name, asset type and condition where possible.
Notification paths depend on the operation. Some teams need email for daily review. Others need SMS or app alerts for urgent faults. Escalation rules can help when an alarm remains active.
Match the design to the application
The same architecture can support many industries, but each application has different priorities.
Water networks
Water networks often need district flow monitoring, leak detection support and night flow analysis. LoRaWAN can suit spread-out low-power assets, while 4G may suit major chambers and pressure management sites.
Key values include flow rate, daily total, minimum night flow and device health.
Irrigation
Irrigation systems need visibility across pumps, channels, offtakes and storage points. Solar power is common, so power use matters. LoRaWAN can work well when a farm or district has gateway coverage.
Useful alarms include high flow, no flow during irrigation, and unexpected flow outside the watering schedule.
Mining
Mining sites often have long distances, harsh conditions and changing infrastructure. Flow monitoring may cover dewatering, process water, dust suppression and tailings-related water movement. 4G, private LTE or site radio networks may be used depending on coverage.
Enclosures, surge protection, cable protection and mounting quality are especially important.
Pump stations
Pump stations benefit from combining flow with pump status, level and pressure. That gives a clearer picture of performance. If a pump runs but flow stays low, the system can flag a likely mechanical, blockage or valve issue.
4G is common where mains power is available, but LoRaWAN also suits smaller sites.
Remote pipelines
Remote pipelines need early warning of unusual flow patterns and stopped flow. The main challenge is coverage. Some sites may use 4G where available and LoRaWAN where gateways can be installed at high points.
A regular heartbeat signal helps confirm the monitoring system is alive, even when flow is steady.
Wastewater
Wastewater sites need equipment that can handle wet, corrosive and confined environments. Meter selection and installation quality are critical. Telemetry can report flow, surcharge events and abnormal patterns.
Alarm delays and filtering help reduce false alerts caused by turbulent conditions or short-term process changes.
Commission the system in a practical order
A structured commissioning process prevents guesswork. Work from the meter outward.
Confirm the meter reading
Check the local display and make sure the meter is installed in the right direction with the correct pipe conditions.
Test Modbus locally
Read the target registers with a local tool or the RTU configuration page.
Check scaling
Confirm units and totals. Compare platform values with the meter display.
Verify communications
Test LoRaWAN or 4G signal at the final antenna location, not just beside the cabinet.
Confirm ProSight data
Check live values, timestamps, trends and site labels.
Trigger test alarms
Simulate a no-data condition or threshold event where safe to do so.
Document the build
Record wiring, Modbus settings, register maps, SIM details, antenna type and cabinet photos.
Documentation saves time later. When a technician returns in six months, clear records can turn a long fault-finding visit into a short fix.
Common mistakes to avoid
Most remote monitoring problems come from small gaps in design or commissioning.
Avoid these common issues:
Connecting RS485 A and B without checking vendor labelling
Forgetting Modbus parity, stop bit or slave address settings
Sending raw register values without scaling
Installing antennas inside metal cabinets
Using short test periods that miss network dropouts
Setting alarm thresholds too tight
Ignoring power draw on solar and battery sites
Failing to label cables and devices
Building dashboards that show data but not asset status
A flow monitoring system should be treated as an operational asset, not just an electronics project. The goal is reliable visibility over time.
What a successful setup looks like
A good remote flow monitoring system is quiet most of the time. It reports on schedule, shows clear trends, and only raises alarms when something needs attention.
The architecture is straightforward:
Flow Meter → RS485/Modbus → Dragino/RTU → LoRaWAN or 4G → ProSight → Dashboard & Alarms
That path gives field teams a repeatable way to connect meters across water networks, irrigation schemes, mines, pump stations, remote pipelines and wastewater assets. Start with clean Modbus data, choose the right backhaul, build clear dashboards, and tune alarms around real operating conditions. The result is less guesswork, fewer unnecessary site visits, and faster response when flow changes matter.
