Robotic or Manual Picking - The Real Question
The question is rarely framed this way. Most discussions start with a technology: a vendor shows robots, and the warehouse wonders whether it needs them. The more useful order is the reverse. First decide whether the switch from manual picking pays off at all, then talk architecture. This page separates the two. It shows where manual picking stays the right choice, what “robotic” actually means - the term hides three very different approaches - and when the switch becomes economical.
Manual Picking - When It Stays the Right Choice
Manual picking is the baseline every automation has to beat. It works in any building, with any assortment and any order mix. Replace it too early and you lose the exact flexibility that keeps many warehouses competitive.
Under the person-to-goods principle, the picker walks to the goods. A WMS instruction or pick list structures the route, but the work itself happens on foot - with a cart, handheld scanner or voice headset. Aids like pick-by-voice or pick-by-light cut the error rate, but they do not change the core: a large share of the shift is spent walking. In manual shelf-based warehouses, roughly half of working time goes to travel between pick positions, at 10-16 km per shift in larger facilities.
Where Manual Picking Is Strong - and Where It Breaks
Manual picking carries the lowest upfront investment of any picking strategy and the highest flexibility. A new SKU needs a storage slot, not a system change. Volume spikes are absorbed with extra staff, assortment changes cost no re-engineering. Wherever volatility is high and predictability is low, that stays a real advantage - and no automation buys it back cheaply.
The limit is structural: manual picking only scales through headcount. More orders mean more people or longer shifts, and in a tight labor market both get expensive and uncertain. As the warehouse grows, the travel share rises too, because the distance between pick positions widens. Beyond roughly 5,000 picks per day the strategy turns inefficient, long before it becomes operationally impossible. So the real question is: how high is my travel share, and how reliably can I get the staff that linear scaling assumes?
Two points clearly go to the manual side: total economics at very low volumes, where no automation model pays for itself, and immediate readiness with zero lead time. If you need more capacity next week and change sites in six months, extra staff serves you better than any automation.
Most existing warehouses do not sit cleanly on one side. That is exactly what the hybrid path is for.
Productivity and Ergonomics in Practice
Collaborative-picking deployments show the order of magnitude of the productivity jump: GEODIS reports an increase from 100 to 175-200 units per hour, Kenco an increase from 30-40 to 120-150 units per hour. These numbers come from live distribution centers, and they already apply to the variant where the picker keeps walking. True goods-to-person raises the lever again, because the remaining travel disappears too.
Ergonomics matter just as much. A stationary workstation with a fixed reach height noticeably reduces physical strain and opens the pick area to staff who can no longer manage a walking shift for health or age reasons. In a labor market where every available worker counts, that is a value in its own right. Relieving the picking process robotically also lowers dependence on a resource that keeps getting scarcer.
The Hybrid Path - Mixed Operation Instead of Big Bang
Robotic picking is not all-or-nothing. The most expensive mistake in a switch is assuming you have to convert the whole warehouse at once. NEO starts as mixed operation from day one: some shelving rows are served by the robot, the rest keeps running manually. The line between the two is configurable in software and can move as the project progresses, without stopping operations.
This also breaks the peak dilemma. In collaborative picking, robots and pickers work in the same picking process. The fleet covers the base load; in high season, additional pickers top it up flexibly - no fixed capacity that sits idle eleven months a year. And because NEO bills on a pay-per-pick basis, operation delivers hard data for the next expansion stage: real picks instead of projections. Start small, and you build the business case for expansion out of live operation.
More on the pay-per-pick model
NEO in Your Existing Shelving
NEO is true goods-to-person from existing shelving. The robot takes individual bins from your current shelving and brings them to a stationary picking station. The racking stays where it is, no new storage infrastructure is built, and the only structural work is the station itself, with power and data connections.
For existing warehouses with medium volume - up to the system ceiling of 100,000 picks per day - this retrofit in mixed operation is usually the most economical way out of manual picking: no capex, no rebuild, live in 6-8 weeks. NEO handles items up to 380 × 270 × 140 mm and 5 kg, in aisles from 85 cm wide. See how the retrofit works, how NEO picks and how pay-per-pick bills.