EDGE COMPUTING PATTERNS FOR REAL-TIME ORDER FLOW OPTIMIZATION

Abstract

The fast digitalization of Quick-Service Restaurants (QSRs) has naturally resulted in the extensive use of mobile apps, Kiosks, and third-party delivery platforms that are currently contributing up to 80% of sales in the leading chains such as McDonald’s and Domino’s. This change has increased the difficulty in handling high volume, multi-channel order flows, especially when peaks are experienced and at lunchtime, where the average drive-through service time is about 5m43s. This research suggests that edge computing patterns should be used with CI/CD pipelines to solve latency challenges and enhance throughput and order accuracy in QSR settings. The study seeks to create and test edge computing solutions to minimize the order-processing latency and improve operational efficiency with the help of simulated and real-world-inspired workloads on the QSR traffic patterns. The most important indicators, like P50, P95 latency, throughput, and error rates, were gathered to evaluate performance. The findings indicated that edge computing had the potential to minimize P95 latency up to 30 to 60%, as well as enhance throughput up to 10 to 20% during peak demand, without eliminating or decreasing order accuracy. The improvements are indicative of huge possibilities of customer wait reduction, resilient in the face of network problems, and optimality in operations of large-scale QSRs such as McDonald’s and Starbucks. Future research topics encompass incorporating 5G and MEC to achieve ultra-low-latency performance and the use of artificial intelligence (AI) to improve demand forecasting in order to augment QSR operations.