Research: Warehouse Workers Do Best when They Switch Between Co-Bots

Why It Matters

Dynamic human‑robot collaboration can significantly lift order‑picking throughput, giving e‑commerce warehouses a competitive edge as they scale automation investments.

Key Takeaways

• Swarm policy boosts picker throughput versus fixed robot pairing
• Performance gains rise as AMR speed and quantity increase
• System-directed approach works only when robot and picker speeds match
• Study simulated 12,000+ scenarios, confirming flexibility's impact on throughput
• E‑commerce growth drives warehouses to adopt autonomous mobile robots

Pulse Analysis

E‑commerce’s explosive growth has forced distribution centers to rethink traditional labor models. Companies are increasingly deploying autonomous mobile robots (AMRs) to assist human pickers, aiming to reduce labor costs and accelerate order fulfillment. While the technology promises higher speed and consistency, the human element remains critical for tasks that require judgment and adaptability. Understanding how best to integrate workers with these machines is now a strategic priority for logistics leaders seeking to maximize throughput without sacrificing flexibility.

The Erasmus University research, published in Transportation Science, evaluated more than 12,000 warehouse configurations using analytical modeling and simulation. It compared a "swarm" policy—where pickers rotate among several AMRs during a shift—with a system‑directed policy that locks a picker to a single robot for an entire order. Results showed the swarm approach consistently delivered higher throughput, and the advantage grew as robot speed and fleet size increased. Only when robot and picker speeds were closely matched, orders were large, and AMR numbers limited did the system‑directed model hold a marginal edge.

For warehouse operators, the takeaway is clear: flexibility in human‑robot pairing can unlock substantial productivity gains. Managers should design task‑allocation algorithms that allow pickers to switch between robots based on real‑time demand and robot availability. Investing in software platforms that support dynamic assignment will become as important as the hardware itself. As AMR technology continues to evolve, the strategic focus will shift from simply adding more robots to orchestrating collaborative workflows that leverage both human intuition and robotic speed.