Here's the conclusion up front: the cheapest component in a pick-to-light system is the battery, and it is also the component most likely to shut the whole pick line down. Pick to light battery maintenance never makes it onto the operations checklist, and sooner or later a warehouse pays for that omission in peak season — not in battery money, but in mispicks, rework, and stopped lines. I have watched this exact failure play out three times.

I've Seen This Hole Three Times

The first time was October 2024, at a 3PL warehouse in East China. Two weeks before peak, 40-plus of 400 light modules in the pick zone went dark. With no lights, pickers fell back on memory and paper lists. The mispick rate climbed from 0.3% to over 2% in two weeks. Rework alone ate three people full-time, two customer complaints came in, and the warehouse manager called me at midnight asking whether the WMS was broken. The WMS was fine. The batteries were not — that whole batch of modules had been installed two years earlier, and their batteries hit end of life together.

The second time was worse because it was my own project. In 2023 I was leading a WMS implementation, and the client had bought the pick-to-light hardware separately. On go-live eve my checklist was all green: interfaces, locations, permissions, printers, exception flows. Nobody asked one question: when do these light modules need new batteries? Eight months after go-live, modules in one zone started dimming and flickering, and pickers reported slow confirm buttons. By the time we traced it to batteries, outbound had been affected for three days. The client's exact words: "Didn't you test all of this before go-live?" We had tested it. We just never wrote battery degradation into the maintenance plan.

The third time was the cheapest failure and the most annoying. An e-commerce warehouse had dead modules, and pickers worked around them by writing the dark locations in a notebook and picking from experience. Three months later the cycle-count variance had more than doubled. The batteries they saved cost less than a thousand yuan. The inventory write-off would have bought batteries for years.

Batteries Don't Die One by One. They Collapse as a Batch

Wireless light modules have one nasty characteristic: modules bought and installed in the same batch have nearly identical battery degradation curves. They don't fail one today and one tomorrow to give you warning. They collapse together within a few weeks.

There are four warning signs, in the order they appear. First, the light dims — nearly invisible in bright daylight. Second, it flickers, and everyone assumes it is a loose connection. Third, response slows: you press confirm and wait half a second for the light to go out. Fourth, dropped signals: the WMS sends the light-up command, the module never responds, the system thinks the light is on, and the picker thinks there is no task.

My rule: start swapping the moment you see the first sign — the first two get dismissed as minor issues, and there is no "good enough" with light brightness. If the picker cannot see the light clearly, the system has effectively issued no instruction.

And here is the counterintuitive part: the peak failure period is not peak season itself. It is the month before. As pre-peak volume ramps up, modules light up more often, and end-of-life batteries fail in a cluster. If you wait until peak day to swap batteries, you are swapping them on a live picking line, and the price is completely different.

The Math: Battery Money Versus Downtime Money

For a pick line with 400 light modules, here is the math from projects I have worked on:

Item Cost (approx.) Notes
One battery A few dollars Price varies several-fold by model; check the vendor spec sheet
One full-line battery swap Roughly one to two thousand dollars 400 modules, about half a day of labor
One year of spare stock A few hundred dollars 15% of total module count
Rework from one mass battery failure Several thousand dollars and up Mispicks, re-checks, re-picks
Half a day of stopped lines in peak Far more than the annual battery budget Late shipments, penalties, and lost customers on top

No ROI model needed: the entire annual battery budget is smaller than the rework cost of a single mass failure. Pick to light battery maintenance is one of those rare cases where the cheap preventive spend is obviously right — the only correct way to save money on batteries is to spend a little on schedule, instead of waiting for it to become a lot.

Pick-to-Light Battery Maintenance in Four Steps

Step one, build a register. Module number, install date, battery model, last swap date — one table. No register means swapping on gut feel. If the WMS doesn't track it, start with a spreadsheet — don't wait for a system enhancement.

Step two, swap in batches, never the whole line at once. Split modules into three groups by install batch and swap one group every four months. That way at most a third of your modules are ever near end of life, and a whole-batch collapse becomes impossible. Put it in the monthly plan, same day as cycle counting.

Step three, run a ten-minute weekly inspection. Check ten modules for brightness. Look for flicker. Press three confirm buttons and check the light goes out instantly. Log any problem module and swap it that week. Assign the inspection to one named person. The day it becomes "whoever is free," the checklist is dead paper.

Step four, keep spares at 15% of the total module count — batteries and spare modules both. Reorder when stock drops below 10%. Don't wait until you run out. Don't stockpile two years' worth, either; batteries have a shelf life. Keep six months and rotate.

Those four steps cost under half a person-day a month. Add one line to the go-live eve checklist — "battery register built, first swap date scheduled" — and the loop is closed.

Wired Power or Battery Power? Decide Before You Buy

New installs often ask whether wired power solves this once and for all. It depends.

Factor Battery-powered Wired
Upfront cost Low, no cabling work High, cabling and installation dominate
Layout flexibility High — move the rack, the module moves with it Low — every layout change means re-cabling
Day-to-day maintenance Battery swaps and a register to manage Almost nothing to manage on power
Failure mode Batteries degrade together, but predictably A power outage takes down the whole zone; needs a UPS
Best fit Layouts that change often, leased sites, light picking Stable layout for five-plus years, owned sites, heavy picking

The trade-off is clean. Battery power buys flexibility and charges you ongoing maintenance attention. Wired power buys peace of mind and charges you upfront cost and layout rigidity. Leased site, layout changing twice a year? Don't go wired. Owned site, layout untouched for five years? Wired wins. The worst choice is picking battery power and then maintaining it like wired power — install it and forget it. That buys the disadvantages of both options.

The Usual Ways This Goes Wrong

First, mixing batteries. Different brands or batches scramble the degradation curve and kill your register's predictive value. One model, one supplier, no exceptions.

Second, only swapping modules that are fully dead. Waiting for a picker report means you outsourced inspection to your error rate. Swap dim and flickering modules the same week. Don't wait for them to go dark.

Third, swapping the battery and not logging it. Skip the log and next time you're guessing the date again — the easiest step to skip, and the one you must never skip.

Fourth, skipping the pre-peak inspection. One month before peak, walk the line, check every module's brightness, swap anything near end of life early. There is no "quick battery change" on a peak-season day. There is only a stopped line.

It comes down to one line on a maintenance checklist. Write it there and the system works as designed. Leave it off and expensive automation loses to a battery worth a few dollars. One thing to do tomorrow morning: walk the pick zone and look for a flickering light or a dim one. If you find one, start with step one and build the register.

By Lin Xiao, WMS Implementation Consultant