
Planning Your IoT-Driven Workplace Safety Strategy
Industrial IoT (Internet of Things) technology works by connecting smart devices to local networks, which gather data about their surroundings. This data is
Use Cases · IoT Sensors
A solar array can lose output for weeks before anyone notices — a shaded string, a failing inverter, a box quietly underperforming. Collie’s dashboard tracks production down to the individual panel, so a dip gets caught the day it starts, not the quarter it shows up in the bill.
Why it matters
Solar underperformance rarely fails loudly. It just produces a little less every day, until someone finally compares the bill to the forecast.
Dust, shading, and ageing cells reduce output gradually — nothing dramatic enough to trigger a manual inspection on its own.
Without panel-level data, a site only sees the combined number — never which specific panel or box is dragging the average down.
Sending someone to physically inspect every panel across a large roof or field is slow, and it only checks whatever day someone shows up.
What’s inside
The dashboard zooms from a single total-output number all the way down to the specific panel that needs attention.
Live and historical kWh output, peak power, and efficiency, tracked continuously against expected generation.
Every panel or string in the array, listed with its current status — producing normally, underperforming, or offline.
Drill into a single combiner box or panel to see its voltage, current, and yield against its own historical baseline.
When a panel or box drifts below its expected yield, an alert is raised automatically — no need to wait for the monthly review.
How it flows
Every panel and box reports its output continuously, feeding one live production number for the whole site.
Each unit’s output is measured against its own expected yield — not a generic average that hides individual problems.
A drop beyond the expected range flags that specific panel or box, so maintenance knows exactly where to look.
A rooftop panel in full sun and one partly shaded by a vent shouldn’t be judged against the same number. Collie compares each panel or box to its own expected output, so an alert always points to a real, specific problem — not just normal variation across the array.
MOD.03 · Also on the fleet
For large rooftop or ground-mounted arrays, a Collie robot can carry the same vision module used for site surveillance to visually inspect flagged panels up close — cutting down on manual roof access for routine checks.
Where it’s deployed
Rooftop and yard installations feeding production floors, tracked against facility energy demand.
Backup and supplementary solar capacity monitored for reliability alongside critical loads.
Corporate rooftop solar tied to sustainability reporting and energy cost tracking.
Large flat-roof arrays where panel-level detail matters most, given the sheer number of units.
It monitors panel-level performance data such as output, efficiency, and fault conditions, so underperforming or damaged panels can be identified without manual inspection.
The system compares live output against expected performance and flags anomalies as they occur, rather than only surfacing them at the next scheduled maintenance visit.
Yes – panel data from multiple installations feeds into the same Collie dashboard, so a facilities team can track several sites without switching systems.
Historical performance trends make it easier to prioritise which panels or strings need attention first, turning maintenance into a data-driven schedule instead of a fixed calendar routine.
Solar Panel Monitoring is built to pull readings from panel-level sensors and standard monitoring hardware into the Collie dashboard – get in touch to confirm compatibility with your specific inverter setup.

Industrial IoT (Internet of Things) technology works by connecting smart devices to local networks, which gather data about their surroundings. This data is
Tell us about your installation — we’ll show you what panel-level visibility and efficiency alerts would look like on your site.