.png)
Complete Industrial Monitoring System Architecture with ProSense from Sensors to Remote Control
Industrial monitoring fails when it is treated as a screen at the end of the process. A reliable system starts much earlier, at the measurement point, and depends on every layer between the sensor, controller, network, software platform, operator dashboard and remote control path.
A complete architecture needs to answer practical plant questions:
Is the value being measured accurately?
Is the signal clean enough to trust?
Can the controller act on it safely?
Can the data reach the right people and systems?
Are alarms meaningful, traceable and timely?
Can operators monitor, report and control assets without increasing risk?
ProSense can support this at different levels. Some facilities only need a single pressure transmitter, level switch or temperature probe. Others need a connected monitoring system designed from field instrumentation through to ProSight software, dashboards, alarms, reports and remote control.
A complete architecture starts with the measurement layer
The instrumentation layer is where the system touches the physical process. If this layer is weak, the rest of the architecture only moves poor data faster.
Common measurement points include:
Pressure in vessels, filters, hydraulic lines and compressed air systems
Temperature in tanks, ovens, chillers, pumps and bearings
Flow in water, chemical, gas or steam lines
Level in silos, tanks, pits and reservoirs
Vibration on motors, gearboxes and rotating equipment
Conductivity, pH or other quality-related process variables
Position, speed and proximity on machinery
The right instrument depends on the media, process range, required accuracy, mounting conditions and environmental exposure. A washdown food facility, a mining pump station and a water treatment plant may all measure pressure, but they often need different wetted materials, ingress ratings, connection types and calibration ranges.
ProSense can supply individual instruments for existing systems, such as a replacement transmitter or a new sensor for an added measurement point. The same instrumentation can also form the foundation of a larger connected system when paired with signal conditioning, controllers, communications and ProSight.
Sensor selection should match the process, not just the tag
A sensor tag might say `LT-204`, but the design work behind that tag matters. Key selection factors include:
Design factor | Why it matters |
Measurement range | Prevents low resolution at normal operation and avoids saturation at high load |
Output type | Determines compatibility with PLCs, RTUs, IoT nodes and signal conditioners |
Process connection | Affects installation safety, hygiene and serviceability |
Materials | Protects against corrosion, abrasion and contamination |
Environmental rating | Supports operation in wet, dusty, hot or outdoor conditions |
Response time | Controls how quickly the monitoring system sees a change |
Calibration requirements | Maintains confidence in long-term measurement quality |
For many industrial sites, the best result comes from standardising where possible. Fewer sensor families can mean simpler spares, clearer maintenance procedures and easier training.
Signal processing turns raw measurement into usable data
Field signals rarely travel in perfect conditions. Electrical noise, long cable runs, ground loops, switching equipment and harsh environments can all affect signal quality. Signal processing protects the integrity of the measurement before it reaches the control and software layers.
Typical signal processing functions include:
Filtering high-frequency noise
Scaling raw values into engineering units
Isolating signals between field devices and control panels
Converting between output types, such as analogue to digital
Linearising non-linear sensor outputs
Detecting out-of-range or failed signals
Applying damping to unstable readings
The most familiar industrial analogue signals are often 4-20 mA and 0-10 V. Digital options may include pulse outputs, serial communications or Ethernet-based protocols. Each choice affects cable distance, noise immunity, diagnostics and the amount of information available from the device.
A 4-20 mA pressure transmitter, for example, can provide a stable process variable over long distances. A smart digital instrument may provide the process value plus device status, diagnostics and configuration data. The right choice depends on whether the system only needs a value or needs richer device-level information.
PLC, RTU and IoT node integration connects the field to logic
Once signals are clean and scaled, they must connect to a device that can read, process and act on them. In industrial monitoring, that device is usually a PLC, RTU or IoT node.
Each has a different role.
Device type | Best fit | Typical strengths |
PLC | Machines, skids and plant control | Fast logic, deterministic control, strong I/O integration |
RTU | Remote assets and utilities | Wide-area communications, local autonomy, low-power options |
IoT node | Distributed monitoring points | Compact data collection, edge processing, cloud or software integration |
A PLC may control a production line or pumping system while also forwarding live values to ProSight. An RTU may monitor remote tanks, bore pumps or environmental conditions where communications are intermittent. An IoT node may gather data from multiple small assets where a full control panel would be excessive.
The integration layer should handle more than basic data collection. It should also manage:
Engineering unit scaling
Signal validity checks
Local thresholds
Time stamping where needed
Store-and-forward behaviour during network outages
Safe default states for controlled outputs
Clear tag naming that matches the asset structure
This is where system design starts to matter. A facility can buy instruments one at a time, or ProSense can design the full path from measurement through to the software model. That helps reduce mismatched ranges, duplicated tags and unclear ownership between electrical, controls and operations teams.
Network communication carries data without losing context
Industrial monitoring networks move data from controllers and field nodes to software systems. Good network design focuses on availability, security, latency and maintainability.
Common communication paths include:
Ethernet within control panels and plant networks
Serial links for legacy devices and remote instrumentation
Cellular connections for distributed or unmanned sites
Wi-Fi where cabling is difficult and the environment allows it
Fibre for longer distances or electrically noisy areas
VPN access for managed remote support
Protocol choice depends on the installed equipment and performance requirements. A plant network may use industrial Ethernet protocols for controller data, while a remote monitoring node may publish data through MQTT or a similar message-based method. Legacy systems may still rely on Modbus RTU or Modbus TCP.
The goal is not just to “get data online”. The goal is to preserve context. A pressure reading should arrive with a tag, engineering unit, timestamp, status and asset relationship. Without that structure, dashboards become a collection of numbers rather than a working monitoring system.
Security belongs in the architecture from the start
Remote visibility and control add value, but they also expand the attack surface. A sound architecture separates networks, controls access and keeps remote operations auditable.
Good practice includes:
Network segmentation between business systems and control networks
Role-based access in software
Secure remote access rather than open inbound connections
Device hardening and managed credentials
Logged control actions
Defined approval paths for remote operation
Regular backups of controller and software configurations
Security design should match the risk profile of the site. A read-only environmental monitoring station has a different risk level from a chemical dosing skid with remote start and stop capability.
ProSight software provides the operating layer
ProSight sits above the field and control layers to turn live industrial data into a usable operating environment. It can give teams a single place to view assets, trends, alarms, reports and remote control functions, depending on the system design.
At this layer, the architecture changes from signals and registers to assets and decisions. A pump station becomes more than several analogue inputs. It becomes a monitored asset with status, pressure, flow, runtime, alarms, maintenance indicators and historical performance.
A well-configured ProSight system can include:
Live asset status
Process values in engineering units
Trend visualisation
Alarm handling
Event history
User roles and permissions
Scheduled reporting
Remote control functions
Multi-site monitoring
Data export for maintenance or operations analysis
This is also where naming discipline pays off. Tags, assets, dashboards and reports should all follow a consistent structure. For example, a site may group data by region, plant, area, asset and signal. That makes it easier to find equipment, compare similar assets and scale the system later.
Dashboards should show what operators need to decide
Dashboards are often the most visible part of the system, but they should not become crowded mimic panels. A good dashboard shows the current state, highlights exceptions and helps operators move from overview to detail.
Useful dashboard functionality includes:
Site overview maps or asset lists
Live status indicators
Key process values
Colour-coded operating states
Trend panels for selected tags
Asset detail pages
Alarm summaries
Runtime and duty status
Maintenance indicators
Mobile-friendly views for field staff
For example, a water monitoring dashboard might show reservoir level, pump status, flow rate, discharge pressure, power availability and communications status. An asset detail page could then show pump starts, runtime, suction pressure, discharge pressure and recent alarms.
The best dashboards support different levels of attention. A manager may need a high-level view across many sites. A technician may need an asset page with recent events and sensor values. An operator may need a focused panel for current alarms and control actions.
Alarm systems need priority, context and discipline
Alarms should direct attention to abnormal conditions that require a response. They should not act as a second trend screen or a noisy notification feed.
A complete alarm design covers:
Alarm setpoints
Deadbands and delays
Priority levels
Acknowledgement rules
Escalation paths
Notification methods
Alarm shelving or suppression rules where appropriate
Event logging
Response guidance
Alarm priority should reflect consequence and urgency. A high tank level close to overflow is not the same as a sensor maintenance warning. Both matter, but they need different notification behaviour.
ProSight can support alarm visibility and event history as part of the monitoring layer. Depending on the architecture, notifications may be sent to operators, maintenance teams or managers. The design should also avoid alarm floods, especially after power restoration, communications outages or process trips.
A useful alarm record should answer four questions:
What happened?
When did it happen?
Which asset was affected?
Who acknowledged or acted on it?
That history supports incident review, maintenance planning and continuous improvement.
Reporting turns history into engineering evidence
Reports help teams move from live monitoring to longer-term understanding. They can support compliance, maintenance, production review, energy management and asset planning.
Common reporting features include:
Daily, weekly or monthly summaries
Runtime reports
Alarm frequency reports
Process trend exports
Exception reports
Sensor health or communications status reports
Pump, motor or equipment duty reports
Environmental or utility monitoring records
For instance, a maintenance team may use runtime reports to balance pump duty or plan service intervals. Operations may use trend data to compare process stability before and after a change. Management may need summary reports across multiple sites to assess availability and recurring issues.
Good reporting depends on good data structure. Tags need units, assets need names, and events need timestamps. If a system captures everything as flat, unnamed data points, reporting becomes time-consuming and prone to interpretation errors.
Remote control must be designed around safety
Remote control can reduce travel, speed up response and improve asset availability. It can also create risk if added casually. The safest architecture treats remote control as a controlled function, not just a button on a screen.
Remote capabilities may include:
Start and stop commands
Mode selection
Setpoint adjustment
Reset commands
Valve open and close actions
Remote enable and disable
Manual override where suitable
Every remote command should pass through defined logic. The controller should confirm permissives, interlocks and current state before acting. The software should show feedback, not simply send a command and assume success.
A safe remote control scheme often includes:
User authentication
Role-based permissions
Command confirmation
Local lockout or remote enable switches
Interlock checks in the PLC or RTU
Event logging
Clear feedback after command execution
Fail-safe behaviour on communications loss
For example, a remote pump start should check level, valve position, overload status, emergency stop condition and local control mode before the motor starts. ProSight can provide the operator interface, but the control logic should remain in the PLC, RTU or local controller where safety and process rules are enforced.
ProSense can supply a component or design the full connected system
Not every project needs the same level of support. Some sites already have control panels, networks and software in place. They may only need specific instruments, such as pressure transmitters, level sensors, temperature probes or signal converters.
Other sites need a complete system architecture. In that case, ProSense can help define how each layer fits together:
Field instrumentation and sensor selection
Signal processing and panel interfaces
PLC, RTU or IoT node integration
Communications between site and software
ProSight configuration
Dashboards and asset views
Alarm design
Reporting outputs
Remote control functions
Expansion planning for future assets
The value of a connected design is consistency. The sensor range matches the process. The signal type suits the controller. The tag name matches the dashboard. The alarm links to the asset. The report uses the same data model as the live view. Each layer supports the next.
That matters across Australia, where many industrial assets operate across large sites, regional facilities and remote locations. A connected monitoring system can reduce unnecessary site visits, improve response times and give teams better visibility across distributed assets.
A well-designed monitoring system behaves like one system
A complete industrial monitoring system is not a collection of unrelated parts. It is a chain of trust from the sensor tip to the operator action.
The measurement must be suitable. The signal must be clean. The PLC, RTU or IoT node must handle the data correctly. The network must carry it reliably. ProSight must present it with context. Dashboards must make the state clear. Alarms must drive the right response. Reports must preserve useful history. Remote control must remain safe, permissioned and traceable.
ProSense can meet the project where it is: supplying individual instruments for an existing installation, or designing a comprehensive connected system that links instrumentation, control, communications and ProSight into one practical architecture. The strongest systems are built with that full path in mind from the start.
