Apache Kafka Streams and Time‐Series ARIMA Modeling for IoT‐Enabled SME Supply Chain Visibility

To provide small and medium enterprises (SMEs) with real-world, low-cost supply chain visibility, this chapter presents an integrated solution based on Apache Kafka streams and a time-series ARIMA (and hybrid ARIMA+ML) modeling solution. Based on procurement, production, warehousing and distribution, SMEs are increasingly embracing the use of Internet of Things (IoT) sensors for visibility in their operations. The chapter examines design pattern, integration strategies and measures of evaluation used in the implementation of these hybrid stream forecasting systems in resource-constrained SME systems and subjects to enhance responsiveness of the supply chain and ensure a lower operational cost. IoT devices installed on supply chains, RFID readers, temperature/humidity sensors, GPS and machine telematics generate time-series information at ever-changing rates, which is heterogeneous. Simplicity, transparency and operational alignment can help organizations to make IoT telemetry reliable decision-making tools that contribute to making supply chains resilient and competitive.