A self-powered wireless node, built on our patented LARA sensor-fusion technology, measures the true tilt and vibration of each tracker, catches the fault before it costs a harvest, and points a technician to the exact failing unit on a map of the plant.
Single-axis trackers lift energy yield by 25 to 35 percent over fixed panels, which is why they now carry almost all new utility plants in Spain.3 But the tracker is the moving, mechanical, weather-exposed weak point of the plant, and operators are nearly blind to the true physical state of each one.
The defense against the single largest cause of solar insurance loss, hail and wind, is the stow position.8 Yet today nobody independently confirms that a given tracker actually reached its safe stow angle. Across a plant of more than ten thousand trackers, faults are still found by yield shortfall after the fact, or by walking the rows.
This is one tracker row with a TrackerGuard node clamped under the torque tube at the row end, so it never shadows the panels. Drag to orbit. Press a scenario to inject a fault and watch the node detect it from the gap between the true measured tilt and the expected sun angle, plus the vibration trace.
The node is a repackaged LARA sensor. The edge logic is our existing thermal-correction and anomaly-detection stack. The dashboard is our existing BIM digital twin. Little has to be invented, which is why this is fast and low risk.
A small cell with a heat-tolerant supercapacitor or hardened battery. Energy autonomous for years.
Fused low-cost accelerometers for true tilt and vibration below 100 Hz.
reuses LARAThermal correction, sun-angle comparison, vibration features, anomaly check.
reuses Thermo-InclinoCalMeasures the structure directly, so a wrong drive sensor is caught, not trusted.
new IPOne gateway covers thousands of nodes across kilometers of open field.
Time sync, alarm aggregation, buffering when the link drops.
NB-IoT or LTE-M, with a satellite link as backup where there is no coverage.
Encrypted, signed firmware, aligned with EU NIS2 for critical energy assets.
Each node tied to a tracker ID, GPS location, drive, and block.
reuses our twinCross-tracker comparison flags the outliers a single node cannot see alone.
A failing tracker lights up on the map. The technician gets a pin, not a spreadsheet.
Which trackers failed to stow during last night wind. Get a located list.
reuses our interfaceLow-cost MEMS is excellent at gross motion and wind faults, good at backlash, and weak at fine bearing wear, because a slew bearing turns too slowly for classic vibration analysis. We sell anomaly and failure detection plus pointing assurance, not magic.
| What the node measures | Fault it reveals | Confidence with low-cost MEMS |
|---|---|---|
| True tilt versus expected sun angle | Stuck row, drift, miscalibration, failure to stow | High |
| Three-axis vibration during wind | Wind-driven torsional galloping and instability | High |
| Motion during a commanded move | Dead motor, jam, blown driver, dead tracker battery | High |
| Angle versus time on reversal | Backlash and mechanical play | With care |
| Slow change of rest geometry | Foundation settlement, structural fatigue | With care |
| Vibration of the slow slew bearing | Early bearing spalling, fine gear wear | Needs other sensing |
| Atmospheric corrosivity (separate area solution) | Corrosion risk of galvanized steel, a few CORTEX units per zone, not on every tracker | High |
Note on corrosion. The CORTEX corrosion sensor is a separate solution, not part of the low-cost per-tracker node. It is more expensive, so it is deployed sparsely across an area of the plant for environmental corrosion anomaly detection, not on every tracker.
The market is strong at controlling trackers and at photographing modules, but every current approach is blind to the true mechanical state of each tracker. That blind spot is the opening.
Nextracker TrueCapture and NX Navigator, Array DuraTrack and SmarTrack, Soltec Dy-WIND. Strong at yield and wind stow, using the tracker own inclinometer and a shared anemometer.
Blind spot. It trusts the drive own sensor, it is proprietary to that brand, and Array passive trackers carry no structural telemetry at all.
Raptor Maps and peers fly thermal drones to find module hotspots and bad connectors and build a digital twin.
Blind spot. It is periodic, not continuous, and it sees electrical and module faults, not mechanical motion, vibration, or galloping.
Plant software flags trackers whose commanded angle deviates from the fleet, catching bad batteries and faulty gears.
Blind spot. It reasons on the commanded or reported angle, so a drifting controller sensor stays invisible and the structure is never measured.
The head to head. Green is a clear yes, amber is partial, red is no.
| Capability | OEM control | Aerial scan | SCADA analytics | TrackerGuard |
|---|---|---|---|---|
| Independent of the drive sensor | No | Yes | No | Yes |
| Continuous vibration and galloping | No | No | No | Yes |
| Per-tracker true tilt | Partial | No | No | Yes |
| Continuous, not periodic | Yes | No | Yes | Yes |
| Works across brands, retrofit | No | Yes | Partial | Yes |
| Cost to instrument every tracker | High | Medium | Low | Low |
We measure the structure directly and compare the true tilt against the expected backtracking and stow angle, so a drifting or wrong controller sensor is caught instead of trusted.
Galloping and gross mechanical faults are caught from the vibration channel on every tracker, continuously, which no control or aerial product offers at this price.
One node fits any tracker make and the legacy installed base, and it is the only structural telemetry option for passive, sensorless trackers.
Built on the LARA multi-MEMS fusion patent, validated on bridges, with thermal-drift correction, so the low cost does not cost accuracy.
The incumbents control the tracker or photograph the modules. We are the only ones that independently and continuously watch the mechanical health of every single tracker, on any brand, cheaply enough to fit them all, and we point a technician to the exact failing unit.
Commercial structural sensors cost thousands of euro per point, which is why no operator instruments every tracker. A real prototype build of the LARA-class board came to about 110 US dollars per unit at a quantity of ten, and the cost is dominated by the MEMS sensor array. At production volume the target is 45 to 70 euro per node, set mainly by the choice of a current-generation MEMS sensor.
Three protected assets, already at TRL 5 to 6 on real structures, transferred to a new field of use.
Spain hosts the largest tracker plants in Europe and a domestic tracker industry, our natural first partner and customer.
Take the validated sensor from one tracker, to one block, to a full commercial plant with an operator partner.
Industrial partner for build and firmware, UPC for the sensor and AI, UCLM for the digital twin, an operator for the site.
Prevented wind and hail failures, recovered yield from corrected pointing, and far fewer truck rolls across huge sites.
Plant and market figures are from public sources. Sensor figures and the commercial-cost baseline are from our internal SAMPIAS proposal. Component prices are volume targets, not quotes.