At a logistics hub in eastern Thailand, the air changes noticeably once 11 p.m. passes. The clatter of forklifts thins out, replaced by the low hum of several picking robots moving between the racks. A lean crew of human workers stays on, focused on the irregular-shaped packages robots struggle with and the items the system fails to recognize.
According to the site manager, this "overnight collaboration" model has really only taken shape over the past year. Overnight shifts used to run with the same staffing levels as daytime, but as hiring dedicated night-shift workers grew harder each year, reliance on robots had to increase.
Working out the rules of coexistence
What's notable is that the site doesn't simply divide work areas between robots and people — it swaps their roles depending on the time of day. During the day, people lead and robots assist; overnight, the roles flip, with people acting more as the robots' "exception handlers."
Do the people move to match the robots, or do the robots move to match the people? It feels like we're redrawing that line every few hours. — a site manager
What makes the handoff work isn't physical barriers like safety fencing — it's a real-time dashboard that visualizes where everything is. The robots track which aisles have people in them and automatically adjust their routes accordingly.
An unexpected side effect: lower turnover
One side effect the company didn't originally anticipate is lower turnover among overnight staff. With robots taking on repetitive, simple tasks, the balance between workload and discretion in the human roles shifted — and that appears to be a key factor.
At the same time, adding too many robots creates new problems of its own, like aisle congestion and competition for charging stations. The search for the "right ratio" of people to robots is still very much a work in progress.