Salman Shikalgar

Project: Raw Material Supply Chain Optimisation & Logistics Analytics.

Project Overview: Developed a scalable data pipeline to optimise the transportation and delivery of industrial raw materials for automotive manufacturing clients. The project focused on reducing lead-time delays, operational costs, and improving route efficiency through data-driven methods and automated status reporting.

1. Situation and Task (The Challenge).

• Managing large-scale, unstructured logistics data for automotive raw materials was leading to supply chain bottlenecks.
• Goal: To streamline inventory management and reduce operational costs by identifying inefficiencies in transit routes and fuel consumption.

2. Action (The Technical Solution)

• Data Orchestration & Storage: Utilised Azure Data Factory for end-to-end pipeline orchestration and Azure Blob Storage for managing large-scale unstructured logistics data.
• Data Ingestion: Engineered modular Python ingestion scripts to process diverse datasets, including transit times and fuel consumption metrics.
• Data Transformation & Analysis: Designed advanced SQL analytical queries to identify specific bottlenecks in the raw material supply chain routes.

3. Result (The Business Impact)

• Efficiency: Optimized the transportation of industrial raw materials, leading to a significant reduction in lead-time delays.
• Cost Reduction: Lowered operational costs through better visibility into fuel consumption and route bottlenecks.
• Stakeholder Value: Streamlined inventory management and improved customer satisfaction for major manufacturing clients.

Project:Industrial (IoT and Machine Breakdown Analytics)

Project Overview: Developed an end-to-end data pipeline to monitor shop-floor machinery and predict potential breakdowns, significantly reducing unplanned downtime in a manufacturing environment. The system calculated key metrics like Overall Equipment Effectiveness (OEE) and Mean Time Between Failures (MTBF) to enable proactive maintenance.

1. Situation & Task (The Challenge)

• Context: High levels of unplanned downtime on the shop floor were impacting production targets and increasing maintenance costs.
• Goal: Build a data-driven failure forecasting system to identify patterns in machine logs and shift from reactive to predictive maintenance.

2. Action (The Technical Solution)

• Data Ingestion & Orchestration: Utilised Azure Data Factory for end-to-end pipeline orchestration and Azure Blob Storage for scalable data lake management of high-frequency sensor logs.
• Data Engineering (Python): Built modular scripts for data cleaning and feature engineering to isolate specific failure patterns from raw machine data.
• Advanced Analytics (SQL): Developed complex analytical queries to aggregate downtime duration and categorise root causes (e.g., mechanical vs. electrical) from relational databases.

3. Result (The Business Impact)

• Operational Visibility: Provided stakeholders with actionable insights through data-driven failure forecasting, allowing for better-informed scheduling.
• Efficiency: Reduced unplanned downtime by implementing proactive maintenance based on calculated OEE and MTBF metrics.
• Scalability: Established a scalable data lake architecture capable of handling increasing volumes of IIoT sensor data.