Strategies for improving supply chain efficiency and economic output based on deep learning and the internet of things in the context of digital transformation

With the accelerated Digital Transformation (DT), Supply Chain Efficiency (SCE) and optimal economic output has become key to the competitive and sustainable growth of an enterprise. Traditional models are usually weak in terms of inaccuracy in forecasting, slow decision-making and inefficiency in logistics. To address these limitations, the research presents an intelligent framework that leverages a novel Boosted Bat Algorithm-Driven Convolutional Multilayer Network (BBA-CMNet) and the Internet of Things (IoT) to modernize traditional Supply Chain performance (SCP) systems. IoT devices, including warehouse sensors, Radio Frequency Identification (RFID) systems, and surveillance cameras, collect continuous streams of structured and unstructured data. Preprocessing involves handling Missing Values (MVs) and applying Z-score normalization to standardize the data. Principal Component Analysis (PCA) is then used for dimensionality reduction as a feature extraction. BBA-CMNet combines Convolutional Neural Networks (CNNs) for accurate classification of warehouse goods and a Multilayer Perceptron (MLP) for forecasting short-term demand using historical trends such as sales and stock levels. The BBA is utilized to optimise the hyperparameters of the model and the dynamic adjustment of logistics paths at the lowest cost in terms of money and time. The experimental findings indicate that accuracy (98.66%), recall, f1-score and precision improve. The BBA-CMNet framework, therefore, provides an AI-centric, scalable framework that does not only improve SC responsiveness and economic output but goals of digital and environmental sustainability.