Resource-Performance Trade-offs in Intra-Aisle Robotic Picking: A Discrete-Event Simulation Comparison of Shuttle Configurations With and Without an Onboard Tote Lift

Mohsen Reda Mahmoud MoursiarXiv 2026 · arXiv

TL;DR

This paper quantifies the engineering trade-off between hardware costs and picking speed in automated warehouses. By replacing two level-dedicated shuttles with a single shuttle equipped with an onboard tote lift, operators can reduce shuttle hardware by 50% at the cost of a roughly 10–15% increase in picking cycle times.

What Problem Does This Paper Address?

In the world of automated warehousing, shuttle-based storage and retrieval systems (SBS/RS) are the workhorses of high-speed order fulfillment. Traditionally, these systems use "level-captive" shuttles—one dedicated robot for every level of the storage rack. While fast, this is hardware-intensive.

A potential alternative is to consolidate resources: using one shuttle to serve two adjacent levels by equipping it with an onboard tote lift. This lift allows the robot to stay on its tracks while raising or lowering the order container to reach items on the level above. The central question the author, Mohsen Moursi, tackles is: Exactly how much performance do we sacrifice when we cut our shuttle count in half? This isn't just about adding a five-second lift delay; it’s about how changing the entire configuration—from task generation to routing logic—impacts the efficiency of the picking cycle.

Methods

The study employs Discrete-Event Simulation (DES) using Siemens Tecnomatix Plant Simulation. The researcher compared two distinct setups:

  • Configuration A (Baseline): Two shuttles, each captive to one of two levels.
  • Configuration B (Consolidated): One shuttle serving both levels using an onboard tote lift (5.0s service time).

The simulation modeled a 30-meter rack with 60 columns and two storage levels. To ensure statistical rigor, the author ran 30 experiments (15 different task sizes ranging from 1 to 15 picking locations, across both configurations) with 10 replications each, totaling 300 observations. The results were analyzed using two-sided Welch t-tests with Holm adjustment to account for multiple comparisons.

Figure 1. Compared shuttle configurations. The onboard tote lift changes level coverage while the shuttle itself remains on its assigned level.

Datasets

The study does not use a standard public dataset but rather a custom-generated benchmark dataset derived from the simulation model.

  • Task Generation: Locations were sampled without replacement from the eligible storage levels.
  • Output: A cleaned 300-row replication-level dataset recording cycle times and item rates.
  • Parameters: Fixed inputs included a shuttle speed of 2.5 m/s, acceleration of 1.5 m/s², and a 10-second pick time per item.

Key Insights and Ideas

  • The "Hardware vs. Time" Ratio: The most striking insight is the disproportionate trade-off. Reducing the shuttle count by 50% (from two to one) only resulted in a ~11–15% increase in cycle time. For many operations, the capital expenditure savings of fewer robots might far outweigh this moderate performance hit.
  • Configuration-Level Impact: The performance drop isn't just a "lift penalty." Because the single shuttle now samples from a larger pool of locations (two levels instead of one), the travel patterns and routing logic change entirely.
  • The Batching Effect: As the number of items picked per cycle ($n$) increases, the relative performance penalty slightly decreases. This is because the fixed costs of the cycle (starting and returning to the home position) are spread across more picks.
  • Statistical Significance: Every single tested task size showed a statistically significant difference ($p < 0.001$), confirming that the onboard lift configuration is consistently slower across all operational scales.

Key Results

The onboard-tote-lift configuration resulted in longer mean cycle times at every tested task size.

  • Cycle Time: At the maximum tested task size ($n=15$ locations), the mean cycle time rose from 299.477s (two shuttles) to 333.231s (one shuttle with lift)—an 11.27% increase.
  • Item Throughput: The derived item rate dropped from 180.314 items/h to 162.050 items/h at $n=15$.
  • Trend: As seen in Figure 2, the gap between the two configurations persists and grows in absolute terms as the number of picking locations increases.
Figure 2. Mean cycle time by shuttle configuration and number of visited picking locations.
Figure 4. Cycle-time-derived item rate (items/h) for each configuration: 3600n / scenario-mean cycle time. These values are not measured two-shuttle concurrent system capacity.

Limitations and Future Work

The authors are careful to note several boundaries to this study:

  1. Concurrent Capacity: The study measures per-cycle performance. It does not account for the fact that two level-captive shuttles can work simultaneously on different orders, whereas one shared shuttle cannot. Therefore, the system-level throughput gap might be larger than the cycle-time gap suggests.
  2. Reliability and Maintenance: A single shuttle serving two levels creates a single point of failure. If that one shuttle breaks, both levels are inaccessible.
  3. Economic Factors: The study quantifies time but not cost. Future work is needed to integrate equipment pricing, energy consumption, and maintenance costs to find the true "cost-optimal" design.
  4. Scope: The results are specific to the 30m rack benchmark and the 5-second lift time; different geometries or faster lift mechanisms would yield different trade-offs.

Summary updated

Abstract

This study compares two implemented intra-aisle robotic-picking configurations through discrete-event simulation. Without an onboard tote lift, two level-captive shuttles serve the two tested levels; with the lift, one shuttle remains on its horizontal level and its onboard mechanism moves the order tote to allow picks from the adjacent upper level. Thirty configuration-task-size experiments (n = 1-15 locations; ten replications each) yield 300 replication-level cycle-time observations. Two-sided Welch tests are applied to the replication means, with Holm adjustment across the 15 comparisons. The onboard-tote-lift configuration has longer mean cycle time at all tested task sizes. At n = 15, the means are 299.477 and 333.231 s/cycle, respectively: a 33.753 s increase (11.27%). The corresponding cycle-time-derived item rates are 180.314 and 162.050 items/h (-10.13%). These rates are calculated as 3600n divided by the scenario-mean cycle time; they are not measurements of the combined concurrent capacity of two level-captive shuttles. Importantly, task locations are sampled from different eligible-level sets in the two configurations, so the result concerns the implemented configuration-and-task-generation packages, not the isolated effect of a five-second tote-lift operation under identical picked locations. The study quantifies a cycle-performance difference alongside a reduction from two dedicated shuttles to one within the tested two-level block; it does not quantify system-level maximum throughput, investment cost, or reliability.
BibTeX
@misc{moursi2026resource,
  author = {Mohsen Reda Mahmoud Moursi},
  title = {{Resource-Performance Trade-offs in Intra-Aisle Robotic Picking: A Discrete-Event Simulation Comparison of Shuttle Configurations With and Without an Onboard Tote Lift}},
  year = {2026},
  eprint = {2610.34708},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2610.34708},
}
Figure