GridOps Sentinel

Bridging the gap between high-voltage infrastructure and Jira workflows with real-time IoT telemetry and automated incident response for the Nigerian Power Grid.

Comment

Inspiration

During my internship as an electrical/electronic engineering student at the Transmission Company of Nigeria (TCN), I experienced the country’s power system instability and witnessed how Nigeria's National Grid serves 200M people while suffering from frequent grid collapses and total blackouts caused by cascading failures that manual monitoring cannot prevent.

In power grid operations (and similar critical infrastructure like hospital ICUs, manufacturing SCADA systems, or automotive assembly lines), the gap between anomaly detection and human response is where catastrophes happen. Grid operators use sophisticated SCADA telemetry but still rely on manual ticket creation when thresholds breach, introducing 5-15 minute delays that can mean the difference between controlled isolation and cascading failure.

It’s quite obvious that while Jira Service Management is the gold standard for incident response, it's fundamentally disconnected from the physical world. Engineering teams in specialized industries need their digital workflows to listen directly to sensors, removing the latency of human intermediation. GridOps Sentinel solves this: it makes Jira sensor-aware for critical infrastructure operations, starting with power grid voltage monitoring.

What it does

GridOps Sentinel is a critical infrastructure telemetry platform built on Atlassian Forge that bridges the gap between physical systems and digital response workflows:

  1. Real-Time SCADA/IoT Integration Forge Webtriggers act as secure HTTPS endpoints accepting high-frequency telemetry from grid sensors Supports standard industrial protocols via JSON payload mapping Handles voltage, frequency, load, and temperature data (extensible to any SCADA metric)
  2. Grid Health Command Dashboard Custom UI embedded in Jira Service Management Project Sidebar Real-time visualization using high-contrast status indicators (Red/Green visual cues) Dynamic threshold monitoring with historical event logging Inspired by power dispatch control room displays (prioritizing split-second readability)
  3. Autonomous Incident Response When telemetry breaches safety thresholds (e.g., voltage spike >250V indicating potential transformer failure), the system instantly creates prioritized Jira tickets with: Exact timestamp and substation location Voltage delta and breach severity Historical pattern context Suggested isolation protocols based on grid topology

Demo Implementation: Nigerian grid voltage monitoring where millisecond detection of instability triggers automatic incident workflows, eliminating the 5-15 minute manual ticket creation delay. Result: Sub-2-second sensor-to-ticket loop that moves grid operators from reactive crisis response to proactive anomaly management.

How we built it

Forge-Native Architecture: Forge Webtriggers: Secure telemetry ingestion (handles auth, scales automatically) Forge Storage: Persistent sensor history (no external database required) Forge Custom UI (React): Real-time dashboard with @forge/react and @forge/bridge Jira Service Management REST API: Programmatic ticket creation with rich context

Engineering Decisions: Serverless-first design: Leveraged Forge's event-driven architecture to handle burst telemetry without infrastructure management Strict CSP compliance: Moved from local dev tunneling to production deployment to satisfy browser security policies Industrial Data Normalization: SCADA systems use inconsistent protocols. We designed a flexible JSON ingestion schema so the system can accept telemetry from various sensor gateways (Voltage, Frequency, Load) without code changes. Minimal external dependencies: Pure React implementation after UI Kit proved unstable Storage optimization: Used concise field names and FIFO (First-In-First-Out) buffering to work efficiently within Forge's storage limits Simulated Data Source: cURL scripts mimic SCADA sensor payloads (actual grid integration requires utility IT approvals beyond hackathon scope).

Challenges we ran into

  1. Content Security Policy Errors Safari and Chrome blocked our Custom UI with blank screens due to CSP violations. We solved this by: Strictly defining all resources in manifest.yml Moving from localhost tunneling to deployed Forge environment Understanding that browser security models don't negotiate—you conform or you fail
  2. React Error #130 & UI Kit Instability Deprecated UI Kit components caused critical rendering failures. We pivoted to building from React primitives (@forge/react), which gave us full control but required rewriting our entire frontend in 48 hours.
  3. Real-Time Data with Serverless Constraints Balancing live telemetry updates with Forge's invocation-based model required intelligent polling—we implemented state management that updates only when new data arrives, not wasteful constant refresh.
  4. Industrial Data Normalization SCADA systems use inconsistent protocols. We designed flexible JSON schema mapping so the system can adapt to different sensor manufacturers without code changes.

Accomplishments that we're proud of

Sub-2-Second Anomaly-to-Action Pipeline We achieved a near-instantaneous response: physical threshold breach → Jira ticket created with full context → dashboard updated in under 2 seconds. For comparison, manual grid operator workflow takes 5-15 minutes (detect anomaly → call supervisor → create ticket → assign team).

Production-Grade Reliability We deployed a stable system that survived CSP battles, handles high-frequency ingestion, and maintains state across Forge's serverless invocations.

Domain-Authentic UX As an EE student, I've seen actual grid control rooms. The dashboard design mimics the displays grid dispatchers use, prioritizing glanceability and status clarity under stress, using visual ASCII bars rather than complex, slow-loading charts.

Extensible Architecture for Critical Infrastructure While we demonstrate power grid monitoring, the pattern applies to any specialized industry with SCADA/IoT telemetry: Healthcare: ICU vital signs monitoring (heart rate, O2 saturation breaches) Manufacturing: Equipment failure prediction (vibration, temperature anomalies) Automotive: Assembly line defect detection (torque sensors, alignment systems)

What we learned

First Principles Engineering Beats Trial-and-Error Working within Forge's constraints taught us to deeply understand security scopes and platform boundaries before coding. We learned to read manifests like electrical schematics, every permission and resource declaration matters.

Stable MVP > Complex Prototype In hackathons (and real engineering), a reliable core system beats 10 half-built features. We ruthlessly prioritized: perfect the ingestion → storage → visualization → ticket creation loop before attempting predictive analytics or multi-site deployments.

Domain Expertise Is the Moat Generic IoT monitoring apps exist everywhere. Our competitive advantage is the understanding that voltage collapses follow predictable pre-fault signatures, that transformer failures cascade in specific topologies, and that grid operators need different workflows than DevOps teams. Insider knowledge makes the difference.

What's next for GridOps Sentinel

Phase 1: Atlassian Rovo Intelligence Layer Deploy an AI Agent that: Analyzes historical incident patterns (voltage anomalies → outage correlation) Suggests pre-emptive isolation protocols based on grid topology and past events Auto-assigns tickets to engineers with relevant substation expertise Acts as a virtual Principal Grid Engineer providing 24/7 intelligent triage

Phase 2: Multi-Site Grid Management Extend from single substation monitoring to regional grid oversight: Cross-substation correlation (detecting cascading failure risk) Load balancing recommendations during high-demand periods Geographic dashboard showing system-wide health

Phase 3: Specialized Industry Templates Package GridOps as configurable Forge app with presets for: Healthcare: Hospital ICU/OR equipment monitoring Manufacturing: Industrial SCADA integration (PLCs, HMIs) Automotive: Assembly line quality control sensors Oil & Gas: Pipeline pressure/flow monitoring

The Vision: GridOps becomes the Forge infrastructure for critical systems monitoring; the bridge between physical operations and Jira's incident management excellence.

Immediate Next Step: Pilot with Nigerian university power labs (University of Ibadan, Federal University of Technology, Akure) to validate with real grid simulators, then approach Distribution Companies (Discos) for production trials.

Built With

Share this project:

Updates