Implementation of Artificial Intelligence (AI) in Google Workspace-Based Warehouse Information System to Optimize Reverse Logistics Truck Scheduling

Abstract

While cloud-based warehouse management systems (WMS) have been widely investigated, their integration with Generative AI to trigger automated decisions in medical device reverse logistics remains underexplored. This study aims to develop and evaluate a low-code Google Workspace WMS integrated with Google Gemini API to optimize reverse flow screening and Full Truck Load (FTL) fleet scheduling at PT Roche Indonesia's TG (Teluk Naga) Transit Warehouse. Employing a descriptive qualitative case study, data was gathered through in-depth interviews with 3 key informants, observations, and system logs. Findings reveal that automated age-based screening (< 7 years) efficiently eliminates administrative delays for 'Destroy' status instruments. Establishing a 75% volumetric load capacity trigger—for both Colt Diesel Double (15 m3 / 4,000 kg) and Tronton Wingbox (48 m3 / 15,000 kg) fleets—proves operationally justified by providing a 25% void-space buffer and a 3-working-day 3PL pickup lead time. Generative AI integration successfully transforms WMS into a proactive decision-triggering mechanism, while user technology readiness acts as a crucial enabler. This research extends the IS Success Model and Technology Readiness Index while offering actionable insights to eliminate warehouse overcapacity risks.