IQZ Fleet Pulse

Objective

FleetPulse will deliver an end-to-end solution to to ingest multiple data sources through batch and real-time processes. It will enable real-time detection of degrading vehicle health, make predictions on expected failure times, and provide real-time responses to driver safety issues. These features will allow businesses that manage a fleet to:

• Make informed and agile real-time decisions
• Interpret fleet health at-a-glance
• Predict vehicle failures in real-time
• Reduce maintenance costs and vehicle turn-over through proactive repairs
• Evaluate driver behavior & safety

Features

Specifically FleetPulse will deliver on the following features:

Ingestion & Processing

Real-time Processing

• Process video and semi-structured data streams with minimal latency to enable agile decision making.
• Prepare and enrich real-time streams for downstream batch processing and busines reporting.

Batch Data & Medallion Processing

• Unify data from external sources.
• Clean and enrich data for clear indications of fleet health.
• Unify data from unstructured, semi-structured, and structured data sources.

AI/ML

Predict Failures and Repairs

• Predict probable diagnostic codes and recommend proactive vehicle repairs to reduce maintenance costs.
• Alert drivers to degrading vehicle conditions before a failure event to reduce vehicle turn-over.

Real-Time Crisis Response

• Monitor for driver collisions and notify emergency response teams in real-time to prioritize driver health safety in emergencies.

Observability

Trip Scorecard

• Score trips for safety to identify vehicles and drivers with the safest records. Enables proactive driver coaching and informed vehicle purchases.

Fleet Health

• Provide insight into overall fleet health to enable informed decisions on vehicle repairs and replacements.

Real-Time Fleet Status

• Visualize data arriving in real-time to provide an accurate to the moment evaluation of ongoing fleet health.
• Enable real-time monitoring of day-to-day fleet challenges as they evolve.

Architecture Overview

The following Azure and Fabric services make up the architecture of FleetPulse:

Batch

• Spark (PySpark) notebooks
• Azure Blob Storage
• OneLake
• Lakehouse

Real-Time

• Azure EventHubs
• Azure EventStreams
• Azure Blob Storage
• Azure EventHouse & KQL Databases
• Real-Time Dashboards

AI/ML

• Azure AI Foundry

Reporting and Monitoring

• Power BI
• Real-Time Dashboards