Verdigris + Eaton Brightlayer + PredictPulse integration

How does Verdigris integrate with Eaton Brightlayer?

Verdigris is a vendor-neutral validated layer that augments an Eaton estate. The integration is additive. Brightlayer Data Centers continues handling device-level SNMP polling across Eaton UPS, ePDU, Tripp Lite rack PDUs, and switchgear. PredictPulse continues 24/7 remote monitoring with phone-based dispatch on critical alarms. Verdigris EV2 sensors clamp on circuits across the full power path and capture continuous 8 kHz waveform data. Brightlayer telemetry and PredictPulse outputs ingest into Verdigris as supplementary context. As the substrate, Verdigris becomes the system of record. Intelligence flows back into your existing systems via MQTT broker, REST API, Modbus, and BACnet emulation. No Brightlayer dashboard goes dark. No PredictPulse contract gets cancelled.

What Eaton does well

Brightlayer Data Centers is Eaton's software suite for device-level monitoring. It consolidates SNMP polling across Eaton UPS systems, ePDU and Tripp Lite rack PDUs, and a deep catalog of switchgear. Polling cadence is configurable per device. PredictPulse is Eaton's 24/7 remote monitoring service. It is not software an operator runs in-house. It is a per-device subscription where Eaton personnel watch UPS telemetry around the clock, apply physics-based early-warning analytics on aggregated data, and pick up the phone on critical alarms. The LTE modem and data plan come included, and field-service dispatch is integrated into the same contract.

Brightlayer is the software. PredictPulse is the service organization that watches the software and picks up the phone when alerts fire. For an operator running an Eaton-standardized fleet, this is real value. Decades of critical-power deployment history. Established channel partners. A service contract behind every UPS.

Most data centers operating today have Eaton hardware somewhere in the power chain. The question is not whether Brightlayer and PredictPulse are good offerings. They are. The real question is what additional intelligence sits above them across the rest of the estate: switchgear, busway, rack PDUs from other vendors, generators. IEEE 1459 defines the power-quality measurement basis for telling useful capacity from waste.

Why not deploy Eaton Power Xpert at high frequency everywhere?

Eaton Power Xpert Meters can capture transients on threshold crossing, but Brightlayer and PredictPulse ingest aggregated telemetry, not continuous waveform.

Power Xpert capture is event-triggered: sub-cycle data exists at the meter for forensic review after a disturbance, then sits there. Brightlayer Data Centers ingests device-level SNMP polling at second-level cadence. PredictPulse is a 24/7 monitoring service (not in-house software) where Eaton personnel run threshold and physics-based analytics on those aggregated streams. The high-resolution waveform lives on the meter; the analytics layer where rules fire never sees it.

Verdigris's pretrained ML models for rectifier degradation, capacitor ESR drift, and harmonic-content failure signatures, trained on years of multi-site continuous waveform data, don't exist on Brightlayer. Standardizing the entire estate on Eaton just to make the high-frequency path work means vendor lock-in across switchgear, UPS, and PDUs. Verdigris is vendor-neutral across mixed estates: Eaton UPS + Schneider switchgear + Vertiv cooling + non-name-brand PDUs all flow into one canonical electrical model.

Where the architecture limits visibility

Brightlayer is device-level SNMP polling. That's the right scope for a hardware vendor's monitoring software. PredictPulse is a 24/7 service applying threshold and physics-based analytics on aggregated UPS telemetry. Both are designed around the Eaton estate. Two architectural consequences follow.

First, the analytics-layer cadence ceiling. SNMP polling reaches the analytics layer at second-level cadence. Threshold and physics-based services like PredictPulse run on that aggregated stream. Sub-cycle degradation (harmonic drift, capacitor ESR shift, transient distortion) falls between samples and does not survive aggregation. The signature is real on the wire; it sits below the polling resolution that reaches the analytics layer.

Eaton Brightlayer uses device-cadence SNMP polling. Verdigris captures 240,000 samples in that window

Second, vendor scope. Brightlayer monitors Eaton devices. PredictPulse watches Eaton UPS. In a mixed estate (Eaton UPS feeding Schneider switchgear, Tripp Lite racks alongside Vertiv PDUs, ABB or Siemens breakers), the non-Eaton kit falls outside Brightlayer's scope, and PredictPulse dispatch is bounded to the Eaton UPS subscription. The operator's power chain crosses two or three monitoring islands by construction.

Consider the 800+ rectifier deployment documented at T-Mobile. Verdigris detected active equipment degradation in approximately 4 percent of the fleet, around 32 units. Zero standard alarms had fired on those units. The degradation lived in 5th and 7th harmonic ratios and in capacitor ESR drift, distortion patterns governed by IEEE 519 harmonic limits. These are below the alarm thresholds that Brightlayer or PredictPulse are configured to monitor. The threshold was the wrong measurement to watch.

How Verdigris adds value to your Eaton estate

Verdigris EV2 sensors install on circuits across the four zones that matter for critical power: utility edge through main switchgear and UPS (greyspace), row and rack distribution (whitespace), the server power path, and the controls zone. The sensors capture continuous 8 kHz waveform data on Verdigris's own measurement floor, independent of any vendor's polling cadence. Brightlayer telemetry and PredictPulse alarm outputs ingest into Verdigris alongside the waveform data, providing device-level context for what the waveform shows.

As the substrate, Verdigris becomes the system of record. The output is a validated operating envelope per circuit, with IEEE 1459-grounded headroom that lets operators reclaim stranded capacity without violating NFPA 70-aligned safe-operating constraints. Brightlayer and PredictPulse keep doing what they do.

Intelligence flows back out to existing systems through four surfaces:

MQTT broker (Verdigris-hosted, real-time pub/sub) for streaming intelligence to other platforms
REST API for programmatic access to electrical state
Modbus for production-validated OT field-protocol integration
BACnet emulation mode: Verdigris sensors appear as native BACnet devices on the operator's network. Zero integration lift. Your BMS sees Verdigris as just another BACnet device; no BMS config change required.

In a typical mixed estate (Eaton UPS, Schneider or Vertiv switchgear, ABB or Siemens breakers, Honeywell BMS, mixed rack PDUs), Verdigris is the one canonical view of the chain. Brightlayer remains the device-level system for Eaton equipment. Verdigris is the cross-vendor intelligence layer above it.

Customer proof

T-Mobile's Senior Director of Infrastructure and Engineering described the architecture shift:

"While competitors react to failures, we're building proactive intelligence that prevents them. Verdigris Signals gives us visibility into degradation patterns that do not exist in BMS telemetry."

Senior Director, Infrastructure and Engineering, T-Mobile

Across the same T-Mobile portfolio, capacity planning improved measurably once validated operating limits replaced nameplate ratings. T-Mobile's Senior Manager of Innovation, Planning, and Strategy:

"Verdigris has been instrumental in refining our capital planning processes, enabling us to make more informed and strategic investment decisions across our facilities."

John Coster, T-Mobile

Across 100-150 MW of planning scope, stranded capacity dropped from 15-20% to under 5%, with on the order of $100M of projected CapEx deferral.

How it deploys in an Eaton shop

A NeoCloud operator runs Eaton 9395 UPS systems. Brightlayer Data Centers polls the fleet at the device level. A PredictPulse subscription covers each UPS for 24/7 remote monitoring and critical-alarm phone dispatch. The downstream chain is mixed: Schneider switchgear, busway from a third vendor, Tripp Lite racks alongside non-Eaton PDUs.

Verdigris EV2 sensors clamp on UPS input feeds and downstream busway circuits, across vendors. Brightlayer continues polling Eaton kit; PredictPulse continues its watch on UPS telemetry; existing phone-dispatch escalation paths stay intact. Verdigris ingests Brightlayer telemetry and PredictPulse outputs alongside its own 8 kHz capture, and exposes intelligence back via BACnet emulation, so the operator's existing BMS sees Verdigris as just another BACnet device with no config change.

Three weeks in, Verdigris detects 4 of 32 rectifiers on the site showing 5th and 7th harmonic shifts. Capacitor ESR drift indicates end-of-life behavior. No PredictPulse threshold has fired; the values sit well inside configured bands. Service is scheduled during the next planned maintenance window, three weeks out. The 4 rectifiers are replaced before any of them fails in production. Brightlayer logs the routine service. PredictPulse logs the planned shutdown. What changed is the warning window: three weeks of lead time instead of 30 minutes between a threshold crossing and an emergency replacement.

The same equipment, the same Eaton vendor relationship, the same Brightlayer dashboards, and the same PredictPulse service contract, with waveform-physics early warning across the full power path on top.

Customer triggers we hear

"Brightlayer already monitors our PDUs." Brightlayer monitors Eaton ePDU and Tripp Lite rack PDUs at the device level via SNMP polling. That's real coverage and you should keep it. Verdigris EV2 sensors clamp on the circuits feeding those PDUs and capture continuous 8 kHz waveform. The PDU dashboard tells you load. The waveform tells you whether the equipment drawing that load is degrading. Different question, different measurement floor. Brightlayer telemetry ingests into Verdigris for context; Verdigris intelligence flows back out via MQTT, REST, Modbus, and BACnet emulation.

"Our Tripp Lite racks have built-in monitoring." Built-in rack PDU monitoring tells you per-outlet current and aggregate load. It does not tell you whether a rectifier feeding that rack is showing harmonic drift, whether a capacitor is approaching ESR end-of-life, or whether a transient on the busway is propagating downstream. The validated layer sits above the rack-level metric and across the whole power path, not in place of it.

Key takeaways

Verdigris integration with Eaton is additive: Brightlayer software stays, the PredictPulse subscription stays, Verdigris adds the validated layer above both.
Brightlayer is device-level SNMP polling. PredictPulse is a 24/7 remote monitoring service with phone-based dispatch. Both are designed around the Eaton estate; neither sees the rest of the power chain.
Verdigris EV2 sensors capture continuous 8 kHz waveform on Verdigris's own measurement floor across the full power path, vendor-neutral.
Intelligence flows back to existing systems via MQTT, REST, Modbus, and BACnet emulation. BACnet emulation is zero integration lift: Verdigris appears as a native BACnet device with no BMS config change.