Aelix AssetGuard · Asset integrity and field operations
Predict the failure before it predicts your outage.
Aelix AssetGuard turns live MQTT and OPC-UA telemetry into a defensible per-asset health score, with domain AI agents that draft the work order, permit-gated and offline-capable field work, and a scoped contractor gateway. Alongside your existing SAP, IBM Maximo, or Oracle EAM (Enterprise Asset Management), not ripping it out.
Agentic AI · Domain agents for asset integrity · Governed by design
Agents that watch the iron, draft the next move, and keep the safety gate human-controlled.
AssetGuard ships domain AI agents purpose-built for asset integrity and field work, not a generic chatbot. They watch your telemetry, reason over live data, draft the next action, and execute it on approval. They run on Aelix Core's built-in agent and workflow engine, the same platform the whole Aelix suite runs on, inside your single-tenant deployment, under role-based permissions, with data never leaving your boundary.
Agents do the work. People keep the authority.
Predictive Maintenance Agent.
Turns telemetry anomalies into a diagnosis and a drafted work order.
Permit Prep Agent.
Assembles OSHA Confined Space / Hot Work permit packages and gates work-order execution.
Contractor Onboarding Agent.
Automates secure contractor vetting, scoping, and gateway access.
Inspection Scheduling Agent.
Builds risk-based inspection and maintenance schedules.
Safety gate stays human-controlled
The Permit Prep Agent assembles the safety package, but the permit-gate stays human-controlled. No agent overrides it and no unsafe work proceeds. Every agent action is logged to the same append-only audit trail as the rest of AssetGuard: governed, attributable, and reversible.
The problem
Your EAM tells you what happened. It cannot tell you what is about to.
Distribution and transmission operators run aging, capital-intensive iron: transformers, switchgear, turbines, transmission corridors. Under rising load and climate stress. Critical assets still fail with little warning under run-to-failure or fixed-calendar maintenance. A single transformer blowout drives outage cost, emergency-procurement premiums, safety exposure, and, in dry corridors, wildfire ignition liability.
Meanwhile, the field work that keeps those assets alive runs on paper and disconnected tools. Crews lose checklist and inspection data in zero-connectivity sites and re-key it later. Safety permits sit in a binder instead of gating the work order. Third-party crews (a large share of line work) onboard over email and spreadsheets with no enforced document-compliance check.
This is the seam. Your Enterprise Asset Management system is your system of record. It knows what you own, what you spent, and what you closed. It was never built to be your system of foresight: a continuous, multi-sensor health signal per asset, and a field-execution layer that proves safe work was done before it was done.
AssetGuard is built to live in that seam: predict the failure, gate the work, capture the field data even off-grid, and write every action to an append-only audit trail your EAM keeps as the source of truth.
What AssetGuard does
Seven capabilities, one auditable spine.
Predictive per-asset health score
A defensible health score per critical asset, derived from live multi-sensor telemetry (temperature, vibration, acoustic, and dissolved gas analysis (DGA)), not a hardcoded number on a dashboard.
Substation gateways stream telemetry over MQTT (mutual-TLS) and OPC-UA into a time-series store. Scored anomaly events from the edge and continuous-aggregate trends drive the per-asset score, with full drill-down into the underlying sensor channels. Target: predictive coverage of at least 80% of Tier-1 assets within 90 days of onboarding.
Digital asset model with live telemetry overlay
A substation asset model with a live diagnostic overlay, so an engineer can see which unit is drifting and why. Sourced from the real telemetry stream, not a random refresh loop.
A 2D isometric substation model is driven directly by the live telemetry feed and per-asset scores. (Honest scope: this is a digital asset model with a live diagnostic overlay, not a photo-realistic 3D twin and not AR/WebXR. A higher-fidelity WebGL model is on the roadmap, stated plainly so you can plan against it.)
Safe work is permit-gated work
OSHA-grade digital safety permits (Confined Space and Hot Work) that gate work-order execution. An unsigned required permit locks the work order. No unsafe work proceeds, and you can prove it.
Each work order carries its required permits. Atmospheric matrix, checklist, and canvas signature are captured, hashed, and chained into the append-only audit log. Target: 100% of gated work orders executed with all required permits signed, zero ungated executions.
Offline-first field capture
Crews complete checklists, permits, and inspections in zero-connectivity remote sites with no lost data. Work is durably queued on the device and auto-syncs on reconnect.
A service worker plus on-device queue holds field mutations offline. Idempotent sync reconciles them when the network returns. Map-based dispatch shows the asset by GPS with a hand-off to native navigation. Target: at least 99% of queued field mutations sync without loss.
Secure contractor gateway
Third-party crews get a scoped, invited portal that exposes only their assigned work orders and their compliance-document uploads. Nothing else in your environment.
Invited, single-company-scoped access with compliance-document templates and expiry rules. Uploads go to isolated object storage. Every action is role-scoped and audit-logged. Target: compliant contractor-crew activation in under two business days.
Augment your EAM, do not replace it
AssetGuard is designed to integrate with your existing EAM (SAP, IBM Maximo, or Oracle), adding predictive telemetry and field execution on top of the system of record you already run.
An integration registry, connection configuration, and connection test ship today. Per-vendor adapters are rolling out, sequenced to the actual EAM in each deployment. (Honest scope: this is designed to integrate with your EAM via a registry, test-connection, and staged adapters, not full bi-directional sync on day one. We sequence one real vendor adapter per pilot against your live environment.)
Repair-vs-replace decision support
A defensible capital model that crosses cumulative maintenance spend against depreciated book value, so you can time CapEx and justify replacements to regulators.
Depreciation schedule input plus maintenance-cost rollup from work orders produces a crossover view and a flag-to-approve workflow. Target: at least 90% of flagged replacements supported by lifecycle crossover analysis.
Why AssetGuard
Honest scope is the differentiator.
Most asset-intelligence pitches collapse in the proof-of-concept because the demo promised more than the product ships. AssetGuard states GA versus roadmap plainly, on the page.
Predictive per-asset health from multi-sensor telemetry, permit-gating, offline-first capture, scoped contractor gateway, append-only audit, the 2D digital asset model with live overlay, and an EAM integration registry with connection testing.
Per-vendor EAM adapters (sequenced to your environment), a higher-fidelity WebGL asset model, and the broader hardening tracked toward general availability.
We would rather you buy what is real and watch the roadmap land than win a demo we cannot survive in the POC.
Built to augment, not rip-and-replace.
Your EAM stays the system of record. AssetGuard adds the system of foresight on top: predictive health and gated field execution. Designed to integrate with SAP, Maximo, or Oracle rather than displace them.
Part of the single-tenant Aelix platform.
AssetGuard deploys single-tenant, anywhere: on-premises or in your private cloud. Application code is identical across deployments. Only environment configuration changes. It runs on the same secure Aelix foundation as the rest of the portfolio: RBAC, append-only audit, and NERC-CIP-aware evidence underneath every action. Isolation is the feature, not the limitation.
What we can prove, and what we are targeting
Mechanisms today. Targets honestly framed.
We do not have a validated customer outcome to publish yet, so we will not fabricate one. Here is what is mechanism-backed today, and the targets we are building to prove.
Mechanisms you can verify in a demo
Targets we are building to prove
All figures above are TARGETS: design goals, not measured customer outcomes.
Security and compliance posture
Designed to NERC CIP standards (CIP-003-9 enforceable 2026-04-01).
AssetGuard is engineered to support your CIP scoping. Not sold as a vendor certification, because compliance is scoped per utility and per deployment.
Append-only audit trail.
A tamper-evident record of state-changing actions, with 7-year retention to support CIP-008 evidence.
Role-based access control (RBAC).
And access-review surfaces aligned to CIP-003-9's tightening governance, vendor-remote-access, and supply-chain requirements.
Encryption in transit and at rest.
TLS across the application, mutual-TLS to substation MQTT gateways, and encryption at rest (LUKS on-premises / KMS in cloud).
Single-tenant deployment in your environment.
On-premises or private cloud. Availability is designed around your operational requirements, not a shared-cloud SLA. We do not publish an uptime figure we cannot stand behind.
Scoped contractor access.
Third-party crews see only their assigned work, every action audit-logged.
Aelix is not certified to NERC CIP, ISO 27001, or SOC 2 by virtue of these statements. Capabilities are described as engineered mechanisms designed to support your compliance program.
See it for yourself
See the score, the gate, and the offline capture on your assets.
Book a working demo against representative telemetry (labeled demo data), or talk directly to an engineer about your EAM, your CIP scoping, and your deployment environment.