Rela AIRela AI Docs
Condition Monitoring

Integrations

CMMS sync, predictive webhooks, smart assignment and automatic briefings

Predictive Integrations

Predictive integrations wire Rela AI's internal pipeline to external systems: the customer's CMMS, third-party webhook consumers, and automatic daily briefings to the plant lead. Every output of the predictive engine has a real recipient.

What is it for?

  • Sync work orders with the corporate CMMS without double-data-entry.
  • Notify external systems when a critical predictive event fires (webhook out).
  • Auto-assign technicians to the best candidate by 5 objective factors.
  • Push a morning briefing via WhatsApp so the plant lead knows what happened in the last 24 h.

How it works

flowchart LR
  P[Predictive pipeline] --> Evt{Event}
  Evt -->|WO needed| CMMS[CMMS sync<br/>SAP PM / Maximo / custom]
  Evt -->|Needs technician| SA[Smart assignment<br/>5 factors]
  Evt -->|External alert| WH[Webhooks<br/>failure_registered / rul_critical / ...]
  P --> BR[Daily briefing<br/>07:00 local]
  BR --> W[WhatsApp to the lead]

All outputs are fire-and-forget: if the external CMMS is down, the work order is saved internally and the sync retries later.

CMMS sync

Rela connects to existing CMMS systems (SAP PM, Maximo, or custom) to synchronise work orders.

POST /api/v1/cmms/configure
{
  "cmms_type": "sap_pm",
  "base_url": "https://sap-server.company.com/api",
  "auth": {"type": "bearer", "token": "..."},
  "field_mapping": {"title": "wo_title", "priority": "wo_priority"},
  "sync_direction": "bidirectional"
}

Predictive webhooks

External systems can subscribe to predictive events:

  • failure_registered — a new failure was logged.
  • rul_critical — a piece of equipment has under 24 h of useful life left.
  • intervention_recommended — urgent intervention is recommended.
POST /api/v1/predictive-webhooks/subscribe
{
  "url": "https://your-system.com/webhook",
  "events": ["failure_registered", "rul_critical"]
}

Smart assignment

The system finds the best technician based on 5 factors:

  1. Proximity (30%) — distance to the equipment.
  2. Specialty (25%) — match with task type.
  3. Workload (20%) — currently open tasks.
  4. Availability (15%) — current status.
  5. Certifications (10%) — LOTO, etc.

If the AI agent creates a work order without picking a technician, the system auto-assigns the best candidate.

Daily briefing

The system generates an automatic morning summary with:

  • Equipment at risk (critical, high, medium).
  • Tasks completed in the last 24 h.
  • Logged failures.

The briefing can be sent over WhatsApp to plant managers.

Key benefits

  • One predictive engine, multiple output destinations (CMMS, webhooks, WhatsApp).
  • Bi-directional CMMS sync with configurable field_mapping — adapts to corporate vocabulary.
  • Smart assignment with 5 weighted factors eliminates arguments over "whose task this is".
  • Predictive webhooks let external systems integrate without having to poll.
  • Automated daily briefing replaces manual shift reports.

On this page