Optimizing Warehouse Picking Operations Using Autonomous Mobile Robots

Order picking is a labor-intensive operation and contributes significantly to overall operational costs in a warehouse. While prior literature suggests collaborative technologies, including autonomous mobile robots (AMRs), can increase picking efficiency, few studies have attempted to find an optimal ratio between AMR and physical pickers using a free-floating policy.

This research develops a simulation model to evaluate the performance of order picking with varying numbers of AMRs and pickers, considering both traditional and cross-aisled warehouse layouts.

Simulation Approach

The proposed simulated order-picking system is loosely based on a real-world foodservice distribution center where humans and AMRs collaborate in a mixed environment. In this warehouse, an entire day's worth of orders for all the DC's customers are received at a single time, in the morning, and order-picking continues until all orders are exhausted (and brought to the loading docks to be distributed).

The discrete event model assumes a free-floating policy. The AMRs are not assigned to a specific picker but rather meet the order picker at a picking position, wait for the picker to load the item, then autonomously drive to the offload point.

To start, both the order picker and the closest available AMR are dispatched to a random unpicked item. After both resources arrive, they pick the item. The picker is then directed to the next randomly selected order. The AMR carrying the picked order is released and autonomously drives the cargo randomly to one of three loading docks (simulating three separate order aggregating points). Once the AMR downloads its cargo, it is dispatched to meet the following order. If either human or AMR arrives at the random item before its complementary resource, it will wait until the other resource comes before jointly completing the picking process.

By conducting and comparing several experiments, researchers were able to examine the following two questions:

  1. What is the optimal ratio for the picker and AMR fleet? The sizing of these two variables is critical since bottlenecks and idle time are to be avoided to maximize operational efficiency and minimize costs.
  2. What is the optimal warehouse layout? The addition of a cross-aisle (or multiple cross-aisles) may reduce travel time and increase efficiency.

Results

The simulation shows how and when warehouse designs meet business requirements. Warehouse simulation is one way to depict the experimentation phase used to plan a warehouse, increase flexibility, and increase a company's productivity. From the standpoint of a professional practitioner, simulation is more intuitive than an empirical approach because racking, forklifts, pit walls, robots, staff, and shift structure can all be easily analyzed using commercial software such as AnyLogic.

The simulation results indicate that the optimal ratio of AMRs to human pickers is 2:1 in the traditional layout and just a bit higher than 2:1 in the cross-aisle layout. These results indicate how a firm can minimize costs by identifying needed resources of each type to maximize resource utilization. Dynamic variables such as labor and AMR, fixed and operating costs must be considered to minimize costs.

Another interesting finding is that simply adding more resources indiscriminately (both AMR and human pickers) does not necessarily increase performance proportionately to the investment.

Postes connexes