Case Study: How Atlantic Water Solved Their Real-Time Challenges

Context (the problem before Azure):

Atlantic Water (confidential) is one of the largest water supply providers, serving millions of households. They faced three critical problems in their IT systems:

1. Sensor Data Overload 🚨

   • Thousands of IoT sensors (water flow, pressure, contamination detectors) sent continuous data.

   • Old system used APIs + databases, which couldn’t handle real-time ingestion.

   • Result → Delays in detecting pipe bursts or contamination.

2. Scattered Event Notifications 📢

   • When a major incident occurred (e.g., pipe burst), multiple teams (maintenance, customer support, billing, emergency services) needed to be alerted.

   • The old system used custom APIs + polling, which was unreliable and duplicated effort.

3. Business Workflows Not Reliable 💸

   • Monthly billing reminders & outage refunds often got lost in the system.

   • Customer complaints skyrocketed because emails/SMS were missed or sent twice.

✅ The Solution: Event-Driven Architecture with Azure

We (MCN Solutions) redesigned the architecture using Azure Event Grid, Event Hub, and Service Bus.

1️⃣ Event Hub – Real-Time Sensor Monitoring

• All IoT devices were connected to Azure Event Hub.

• Example: 10,000 sensors send pressure & water quality readings per second.

• An Azure Function with EventHubTrigger reads and analyzes data in real time.

👉 Benefit: No more bottleneck. If pressure > 100 PSI, the function flags it instantly.

Problem
"Earlier, our team couldn’t process high-frequency IoT data. By using Event Hub + Azure Functions, we now process millions of events per minute in real-time."

2️⃣ Event Grid – System-Wide Notifications

• When a pipe burst or contamination is detected, the Function publishes an event to Event Grid.

• Event Grid fans out notifications to:

  • Emergency Service (dispatches field crew).

  • Customer Notification Service (sends SMS/email).

  • Billing Service (applies credits/refunds).

👉 Benefit: Every stakeholder is notified instantly, with no polling or tight coupling.

Now:
"We used Event Grid as a lightweight pub/sub backbone. A single contamination event now reaches emergency teams, customer service, and billing — all within seconds."

3️⃣ Service Bus – Reliable Business Workflows

• Customer reminders, outage refunds, and tickets go into Service Bus queues.

• Another Azure Function with ServiceBusTrigger processes them one by one (in order).

• Dead-letter queue ensures failed messages (e.g., unreachable customer) are not lost.

👉 Benefit: Billing and reminders are now guaranteed, ordered, and reliable.

Situation:
"For critical workflows like billing reminders, we relied on Service Bus. Unlike Event Grid, it guarantees delivery, retries, and preserves ordering."

🔥 End-to-End Flow

“Let’s take a real incident: A contamination was detected in Zone-23.”

1. Sensor → Event Hub: IoT device streams ‘contamination detected’.

2. Function → Event Grid: Function processes it and raises an Event Grid event.

3. Event Grid fan-out:

   • Emergency Service dispatched.

   • Customer Notification Service sends alerts.

   • Billing Service applies refund credits.

4. Service Bus Queue: Customer reminders & refund confirmations are queued, processed, and retried if needed.

👉 Result: From detection to customer notification — under 5 seconds.