When push over-delivery looks like good service
Leaving unrequested boxes at the station can feel like good service. Lean still treats that push over-delivery as overproduction.
What push over-delivery is
In lean terms, push releases material from forecast, schedule, or driver judgment before a downstream station signals need. Pull waits for that signal. Classic kanban and just-in-time methods stop upstream supply from filling the line early. On the hall, that failure mode is often logistics, not machining: movers drop extra full units beside the station, park empties wherever space opens, and keep circulating so they will not miss a shout.
Protective early drops raise line-side WIP, shrink usable aisle width, and still starve the station when the wrong mix arrives early. The early drop can look helpful while the floor still pays in space, search time, and a later shortage of the right SKU. That is why lean manufacturing lists overproduction among the wastes under muda, not among customer service wins. Inventory management can still show a healthy part count while the aisle is blocked by the wrong protective stock.
What the habit hides in internal logistics
Under a push habit, loading and unloading often rest on experience. Drivers patrol production looking for needs. They deliver more boxes than requested so they will not be blamed for a shortage. Time burns as empty travel. Floor space burns beside the line. New drivers take longer to learn because the rules live in people's heads, not in claimable jobs. Aisle congestion follows when several trucks keep feeding the same spine with unrequested stock.
Industry e-Kanban and digital material call-off practices replace that protective overstock with a named request. twinzo's AOS (commonly known as FGS) turns the line need into a visible work order under material order automation. Live paths under internal logistics optimization then show whether empty returns and buffer dwell still follow the old push habit.
What pull changed on documented halls
On a documented 30,000 m² (about 323,000 ft²) automotive Tier-2 plant, the operator presses a button for an empty or full box exchange. Production needs appear on a tablet in the forklift in real time. Operator-to-driver phone calls fell by about 90%. Patrolling fell by more than 95%. Parts arrive when claimed, not when a driver guesses. That case publishes no euro investment or saving for the coordination cut itself. The measured change is the call and patrol reduction.
On a documented 55,000 m² (about 592,000 ft²) white-goods plant, the same pull class recovered about 20 minutes of production daily, eliminated micro-stoppages tied to late material, and recorded €5,080,000 (about $5,588,000) increased production value in year one against €100,000 (about $110,000) invested. Named empty call-offs are covered in empties as a first-class call-off. The recovered line minutes are in twenty minutes of production back every day. Quiet daily loss is framed in the small daily loss lean teams keep missing.
What path proof adds after push leaves
Once claimable jobs replace protective drops, a real-time location system (RTLS) shows whether delays and empty corridors actually left. Medium-precision location around 1–3 m (3–10 ft) is enough for aisle, buffer, and bay proof on that question. A documented Tier-2 RTLS case on a 30,000 m² (about 323,000 ft²) hall cut fleet from 36 to 20 forklifts and drivers from 108 to 60, eliminated delivery delays, and saved €960,000 (about $1,056,000) OPEX in year one against €120,000 (about $132,000) invested. Another Tier-2 RTLS case in the same size band lifted productivity from 63% to 75% and freed four FTEs when empty runs became countable.
Over-delivery is a WIP and congestion problem dressed as helpful logistics. Plants also fund location for material, not only movers, because shorter search time lets them run lower WIP without starving the line. Fleet impact is in when buying more forklifts is the wrong first answer. How pull and path share one review is in when pull and path share one hall.
What the twin keeps after the habit change
Closed call-offs and mover paths fill a spatially oriented dataset under the live hall. Live logistics owns whether protective early drops and empty patrol still burn the shift. Continuous improvement owns layout and supermarket sizing from the same place-and-time record. Correlation and causality analytics can ask which early drops still precede aisle blocks or starves. Those three readers sit on equal footing. An operational digital twin that only celebrates fewer phone calls has not finished the job.
Pull without place can still hide residual push in buffer dwell. Place without pull can still dispatch by radio. That incomplete pair is covered in pull and e-Kanban still fail without place.
How teams decide push is no longer acceptable
1. Count extra boxes left at line-side that were not requested on one peak shift - Over-delivery becomes a number, not a story about helpful drivers.
2. Measure empty patrol and phone share on the same loop - High search usually feeds the push habit.
3. Pilot button or scan call-offs for empty and full exchanges before a hall-wide cut - Wire triggers under request triggers in the workflow editor.
4. Keep production waits on the same map as logistics - Line space and starve minutes belong under production monitoring. Capability depth sits on the features overview. How a rollout is priced is on pricing.
Get in touch if you want to walk whether unrequested early drops are still the congestion your fleet keeps paying for.