Introduction
In modern manufacturing, efficient material replenishment is crucial in assembly plants to prevent delays and excessive inventory. Traditional methods often lead to inefficiencies like overstocking and unnecessary travel. Autonomous Mobile Robots (AMRs) offer a solution by automating assembly plants.
This research focuses on AMR-based material replenishment at the Tiger Motors lab (Auburn University), where a Stretch RE1 robot supplies materials for Lego car assembly. Using a multimethod simulation model with discrete-event and agent-based simulation modeling in AnyLogic, the research evaluates different replenishment strategies to minimize travel and optimize efficiency.
Simulation model
The research utilizes a multimethod simulation approach that combines:
- Discrete-event simulation to model the autonomous mobile robots’ movements and task scheduling.
- Agent-based modeling to simulate the behavior of containers and their replenishment needs at assembly plants.
AnyLogic is used as the primary tool for developing and running the model. The autonomous mobile robot, named Stretch RE1, autonomously delivers material containers following different scenarios. The model includes various input parameters (e.g., time to empty a container, trip time, payload capacity) and output measures (e.g., number of shifts required, station idle time, payload utilization).
Three simulation scenarios were tested:
- Carry limit consideration (scenario 1) – the autonomous mobile robot’s objective is to minimize the number of shifts while optimizing the container capacity the robot should carry.
- Full load utilization per shift (scenario 2) – the robot waits to carry a full load before replenishing.
- Optimize replenishment timing (scenario 3) – the AMR replenishes when it reaches full capacity or when a container is nearly empty.
Results
The study demonstrates that the efficiency of autonomous mobile robots in assembly plants depends on selecting the right replenishment strategy. Scenario 1 showed the most effectiveness when the robot has a higher payload capacity, as it reduces the number of trips needed. Scenario 2 proved to be ineffective since it leads to excessive delays and empty stations. Scenario 3 is more suitable when the payload capacity is limited, as it minimizes station idle time.