A warehouse supervisor has robotic vendor demos open on one monitor. On the ground beneath, a picker is working from a paper decide checklist, checking off traces by hand, and keying confirmations right into a terminal in a batch on the finish of the shift, each time there’s a free minute.
The hole between these two footage is often the place the actual return is sitting, and it prices a fraction of what’s on the opposite monitor.
That picker is commonly already carrying a cellphone, a handheld, or a display screen mounted on the cart. It’s simply working one slim app, or nothing related in any respect.
Most of a picker’s shift isn’t spent selecting. Tutorial analysis on order-picking operations constantly places journey between areas at roughly half of complete selecting time, typically extra.
The remaining leaks out in three particular methods, not a common sense of inefficiency. Routes get designed as soon as, when the warehouse carries half the SKUs it does right this moment, and no person’s job is to revamp them because the catalog grows, so pickers preserve strolling a format that not matches how the enterprise sells.
Single-order selecting continues the place quantity has lengthy since justified batching: ten single-item orders picked separately means strolling the identical aisles ten occasions as a substitute of as soon as. And paper confirmations don’t attain the system till hours after the merchandise truly left the shelf.
In that window, the system nonetheless reveals inventory that’s already gone, which is precisely when a second channel sells the identical unit to another person.
A scanning system within the picker’s hand doesn’t substitute selecting. It removes these three particular leaks. Optimized routing sequences every decide so the trail via the warehouse is shortest, not no matter order the printed checklist occurs to run in.
Batch and cluster selecting group a number of orders travelling the identical route into one journey as a substitute of 1 order at a time.
Scan affirmation writes the inventory replace the moment an merchandise is scanned, so the hole between Ventor PRO confirming a decide and the system figuring out about it drops from hours to seconds, which is what truly closes the overselling window described above.
Robots earn their price underneath particular circumstances: excessive, regular order quantity, a fairly steady assortment, and a labor market the place hiring and maintaining pickers is genuinely exhausting.
None of that may be a knock on robots – they’re an actual reply to an actual drawback, as soon as it’s the appropriate drawback. What doesn’t work is automating a course of that’s nonetheless messy, as a result of a robotic doesn’t repair dangerous routing or stale inventory knowledge, it simply repeats no matter course of it inherits, quicker.
A warehouse that’s already scanning, batching, and syncing inventory in actual time is a much better robotics candidate than one nonetheless confirming picks on paper, just because it’s automating a course of that already works. Repair the method first, then automate one which’s already environment friendly.
Earlier than the following robotic demo, it’s price asking a narrower query: of the three leaks above – useless routes, unbatched single-order selecting, and paper confirmations – that are nonetheless costing this warehouse cash right this moment? For lots of operations, closing these is the cheaper step.
It additionally makes the robotic funding repay quicker each time that step comes, as a result of it palms the robotic a course of price automating as a substitute of 1 price fixing.
