Problem:
Inefficiencies in warehouse operations, such as poor resource allocation, bottlenecks, and suboptimal workforce distribution, were resulting in increased costs and delays.
Solution:
Intel developed a simulation-based warehousing model using AnyLogic to forecast the impact of operational changes, optimize workflows, and improve efficiency.
Results:
- Designed and implemented a comprehensive strategy that achieved a 30% increase in warehouse productivity.
- Developed a scenario where 15% more boxes were picked and 40% more boxes were placed.
Introduction: focus on efficiency and scalability
Intel, a global leader in semiconductor manufacturing and technology solutions, continuously seeks innovative ways to optimize its operations. As both a chip designer and a contract manufacturer through Intel Foundry, the company focuses on efficiency and scalability across its business.
Intel’s logistics teams always explore advanced methodologies to enhance productivity and reduce costs. One of their key initiatives involved the development of better strategies for efficient warehouse operations using a simulation model.
If warehouse optimization is a hot topic for you, take a look at our recent blog post.
Problem: aim for efficient warehouse operations
Intel struggled with inefficiencies in its warehouse operations in Chengdu (China), which led to increased costs and delays in order fulfillment. The key challenges included:
- Poor resource allocation meant that warehouse staff and equipment were not used optimally. It resulted in idle time and inefficiencies.
- Bottlenecks in material movement caused frequent obstructions in the flow of goods, slowing down operations and creating delays.
- Imperfect workforce distribution resulted in labor being unevenly assigned (some areas underutilized and others overloaded).
Intel needed a warehousing model that could provide insights into different operational scenarios without disrupting real-time processes. The company required a solution that would allow them to forecast the impact of changes before implementation, improving overall efficiency. By adopting a simulation experience, Intel aimed to refine the processes, eliminate bottlenecks, and achieve truly efficient warehouse operations.
Read also: a case study on an automated warehouse system for a leading footwear chain in North America.
Solution: warehousing model development
After evaluating several potential approaches, the company decided that a simulation-based solution would provide the most accurate and scalable insights.
With technical support from Ormae, Intel selected AnyLogic as their preferred solution. Unlike other software options, AnyLogic enabled them to develop a highly detailed and dynamic warehousing model customized to their specific environment. This was made possible by AnyLogic flexibility and its unique capability to seamlessly integrate agent-based, discrete-event, and system dynamics simulations.
The project began with an extensive data collection phase. Engineers mapped out the warehouse’s layout, capturing the positioning of storage zones, aisles, and docking stations. They recorded real-time operational data, such as worker movement patterns, order processing times, and AGV (Automated Guided Vehicle) usage.
One of the main focus areas of the project was the Autonomous Case Handling Robots (ACR), part of Intel’s warehouse digitalization efforts. Using a simulation model, Intel aimed to analyze their productivity and maximize system efficiency. In the image above, you can see the logic used in the warehousing model to replicate the robots.
Once the simulation was constructed, the team experimented with various what-if scenarios in a risk-free setting. They tested different configurations of workforce deployment, product placement strategies, and automation levels to identify the most efficient warehouse operations structure. The warehousing model provided predictive insights, allowing managers to assess the impact of potential changes before implementation.
The team tested and refined solutions through iterative adjustments. These experiments helped them find the best ways to minimize delays, reduce congestion in high-traffic areas, and maximize resource utilization.
Discover how to design smarter warehouse layouts with AnyLogic simulation modeling.
Results: strategy for a 30% increase in productivity
The implementation of the simulation project yielded significant improvements in warehouse operations.
Key outcomes included the following:
- The warehousing model showed that adjusting outer tote placement could boost productivity by 30%. The different scenarios provided optimized storage arrangements to speed up retrieval and improve workflow.
- The Intel team identified periods of underutilization and peak demand. This helped adjust the number of robots to avoid excess while ensuring efficient warehouse operations.
- The team measured overall warehouse productivity and came up with scenario where 15% more boxes were picked and 40% more boxes were placed, improving efficiency and throughput.
The next steps of the project aim to implement and validate the warehousing model’s recommendations for efficient warehouse operations. The model will be expanded to different warehouse layouts for better adaptability. Finding the right number of robots will help balance cost and productivity.
The case study was presented by Swee Heng Yim and Alankrit Goel from Intel, Noorahamed Gadamphalli from Ormae at the AnyLogic Conference 2024.
The slides are available as a PDF.
