Supply Chain Demand Forecasting & Inventory Optimization
Description
This dataset and analysis were developed as part of a self-initiated Supply Chain Analytics project focused on demand forecasting and inventory optimisation for a retail distribution network. The objective was to demonstrate how open data and free analytical tools can improve forecasting accuracy, reduce stockouts, and optimise reorder policies — without any financial investment. Key Activities • Data cleaning and preparation of order, shipment, and inventory records • Exploratory analysis of sales trends and seasonality • Development of a simple forecasting model (moving average / ARIMA) • Calculation of optimal reorder points and safety-stock levels • Visualisation of performance improvements using Excel and Python dashboards Tools Used : Excel · Google Sheets · Python (pandas, matplotlib) Dataset Source Reference : SMART Supply Chain for Big Data Analysis – Mendeley Outcome : Created a repeatable learning framework for students and professionals interested in practical supply-chain analytics using free resources.
Files
Steps to reproduce
Step 1 – Download Data Download the open dataset SMART Supply Chain for Big Data Analysis from https://data.mendeley.com/datasets/8gx2fvg2k6/3 . The dataset contains order, shipment, customer, and product data useful for supply chain analysis. Step 2 – Import Data Open the dataset in Microsoft Excel, Google Sheets, or Python (pandas). Verify headers and data types (dates, numeric values). Step 3 – Clean Data Remove duplicate and null values. Standardize date formats (e.g., YYYY-MM-DD). Filter columns by relevant variables such as product category, region, and time period. Step 4 – Data Analysis Compute total monthly and category-wise demand. Identify trends and seasonality. Calculate lead time and delivery performance metrics. Step 5 – Forecasting Apply a Simple Moving Average (SMA) or ARIMA model to forecast future monthly demand. Validate the model by comparing actual vs. predicted values. Step 6 – Inventory Optimization Calculate the reorder point and safety stock using the following formulas: Reorder Point = (Average Demand × Lead Time) + Safety Stock Safety Stock = Z × σ × √Lead Time (Z = service level factor, σ = standard deviation of demand) Step 7 – Visualisation Create visual charts for forecast vs. actual demand. Build a simple dashboard in Excel or Google Sheets to display KPIs (inventory level, forecast accuracy, stockout rate). Step 8 – Interpretation Compare pre- and post-optimization results. Summarize improvements in stock availability, forecast accuracy, and service level. Step 9 – Documentation Save the analysis in Excel or Word format and share with the dataset as supplementary material. Tools Used: Excel, Google Sheets, Python (pandas, matplotlib). Outcome: Reproducible analysis workflow for forecasting and optimizing retail supply chain inventory using open data and zero-cost tools.
Institutions
- Indian Institute of Technology Kharagpur
Categories
Funders
- Indian Institute of Technology KharagpurIndiaGrant ID: Independent Research (affiliated with IIT Kharagpur)