AI-powered warehouses: A new era of sustainable inventory management

Editor's Note: The SCM thesis AI-Powered Warehouses: A New Era of Sustainable Inventory Management was authored by Kyungmin Kook and Elisa Ruiz Mugica and supervised by Dr. Josué Velázquez Martínez ([email protected]) and Dr. Miguel Rodríguez García ([email protected]). For more information on this research, please contact the thesis supervisor.

Warehouse operations are often overlooked as contributors to greenhouse gas (GHG) emissions in the logistics sector. Our capstone project set out to measure emissions reductions from improved inventory management. Along with our sponsor company Verity—a provider of AI-powered inventory management systems—we partnered with a global logistics provider to assess the environmental impact of implementing Verity’s indoor drone system in a U.S.-based fulfillment warehouse.

Our research explored how drone-based inventory automation impacts total GHG emissions across Scopes 1, 2, and 3 and which operational levers—labor, equipment, or waste—experience the greatest emissions changes due to automation. (Note: Scope 1 refers to emissions directly produced by an organization; Scope 2 refers to indirect emissions resulting from an organization’s energy use; Scope 3 refers to indirect emissions produced throughout an organization’s value chain.)

Constructing the study

To assess the environmental impact of drone-enabled inventory automation, we developed a mathematical model using real operational data, integrating activity-based emissions modeling with lifecycle assessment (LCA) to estimate emissions changes across Scopes 1, 2, and 3. We collected operational data from both pre- and post-deployment periods, including cycle count records, inventory composition, equipment usage, and staffing levels. When direct data was unavailable, we supplemented it with structured interviews, industry benchmarks, and peer-reviewed literature.

Key operational variables included forklift energy use, inventory write-offs, employee commute distances, and drone charging requirements. Each was mapped to a corresponding emissions scope using standardized emissions factors from sources such as the U.S. Environmental Protection Agency and Verity.

Using the model, we evaluated three main levers to understand the drivers of emissions reduction:

• Inventory accuracy improvements: reduction in inventory write-offs due to more frequent and precise cycle counts enabled by autonomous drone scanning
• Labor efficiency gains: decrease in employee commuting emissions from reduced staffing required for inventory tasks
• Equipment utilization changes: reduction in forklift usage and a decrease in the total number of forklifts required, reducing both energy consumption and lifecycle emissions

We also included lifecycle emissions (Scope 3, LCA) associated with the manufacturing and transport of drones and forklifts. To test robustness, we conducted a sensitivity analysis using three drone coverage scenarios:

• Scenario 1: 64% drone coverage (current)
• Scenario 2: 90% drone coverage (target)
• Scenario 3: 100% drone coverage for scannable locations

The benefits of drone automation

We found that drone automation significantly reduced emissions. At 64% drone coverage, emissions decreased by approximately 49.5% compared to the manual baseline. This reduction was driven by reduced inventory write-offs (Scope 3), reduced forklift usage (Scope 3 and LCA), and reduced commuting by staff (Scope 3). Benefits tapered off beyond 90% drone coverage; increasing drone coverage to 90% led to an additional 33% reduction in emissions relative to the 64% baseline, but further increases yielded diminishing returns, indicating that the majority of benefits are realized before full coverage.

Our findings contribute to a growing body of evidence that warehouse automation, when implemented thoughtfully, can serve as a critical enabler of corporate decarbonization strategies.

SC
MR