Real-time visibility across 14 industrial sites: from blind spots to immediate response

Client Overview

GreenField Industrial is an Australian industrial operations business that manages environmental monitoring equipment across multiple remote sites, including water treatment plants, utility substations, and processing facilities. The operations team is responsible for maintaining equipment health, responding to faults, and producing compliance reports for regulatory bodies. Before this engagement, the team had no unified platform for any of this work.

The Challenge

Device health data existed in isolated systems: some devices reported via SMS, others via proprietary monitoring software that did not integrate with anything else. When a fault occurred, the team found out by phone call, manual site visit, or, in some cases, a regulatory breach notification. Response times were slow. Reporting was done manually from multiple data exports. There was no way to see the status of all sites at once, and no alerting system to catch problems before they became operational failures.

The operations team was flying blind. If a device failed overnight, they often didn't find out until the next morning check-in.

The Solution

PTM Software designed a centralized IoT dashboard that aggregated data from all device types into a single unified interface. The dashboard provided live device status, configurable alert thresholds, and automatic escalation rules, so the right person was notified the moment a reading went out of range. A separate reporting module generated automated compliance summaries, replacing a process that had previously taken the team two days per month to complete manually.

Technical Architecture

Device data was ingested via an MQTT broker, a lightweight messaging protocol well-suited for environments where network connectivity is intermittent or restricted. All time-series data was stored in InfluxDB, which is purpose-built for high-frequency device readings and fast time-based queries. The React dashboard used WebSocket connections to push live updates to the browser without requiring a page refresh. Cloud infrastructure was deployed on a managed platform with redundancy across availability zones to ensure uptime even during regional outages.

• MQTT broker for reliable device data ingestion over low-bandwidth connections
• InfluxDB time-series database for high-frequency telemetry storage and querying
• React dashboard with live WebSocket updates and configurable alert thresholds
• Automated escalation rules with role-based notification routing
• Compliance report generator producing structured PDF outputs
• High-availability cloud infrastructure with zone-level redundancy

How We Delivered It

We began with a two-site pilot. Rather than connecting all 14 sites at once and exposing the team to a large operational risk, we ran the first four weeks on two pilot locations where the team could validate alert thresholds, test the escalation workflow, and identify any gaps before full rollout. This proved critical: the pilot revealed that several alert thresholds needed adjustment for seasonal temperature variation, something that would have generated a flood of false positives across the full site network. By week six, the team had refined the configuration and rollout to all 14 sites was completed by week eight.

Results

Lessons from This Engagement

Phased rollout with a pilot group is not a compromise; it is the responsible way to deploy monitoring infrastructure in a live operational environment. The two-site pilot saved the operations team from a difficult first week of false alarms and allowed the full deployment to land cleanly. In any system where notifications drive real-world action, it is worth investing extra time in validating alert logic before scaling.