The empty-round problem that still sits in most plants
Even on a staffed hall, empty forklift patrol still burns vehicle time and line minutes. Claimable pull is the step that cuts that drain without pretending the plant is already lights-out.
What empty rounds are on a staffed path to darker operations
Empty rounds are avoidable forklift travel spent searching for the next job instead of carrying a load. Claimable pull cuts that waste when drivers take open material jobs from a shared queue instead of hunting for the next radio call. Empty travel is mover distance or time with no productive load. On a dark-factory path, the waste worth cutting first is avoidable search and patrol: circulating for the next radio job, checking boards, or looping a corridor until someone flags the truck. A necessary empty return after a loaded drop is not the same waste. Lean manufacturing names the pattern as transport and motion muda.
Each empty kilometre (empty mile) costs fuel or battery and keeps the vehicle out of loaded service. When the next pallet or kit arrives late, the station takes a micro-stop: a short wait that often never opens a downtime code. Across stations and shifts those waits become real capacity loss. For the same waste at hall scale on a white-goods footprint, see why empty forklift rounds still drain a white-goods plant. This page keeps the dark-factory cost question: what you buy when people still drive and you stop the search.
What one documented AOS spend returned
A documented 55,000 m² (about 592,000 ft²) white-goods plant invested €100,000 (about $110,000) in twinzo logistics management through AOS (commonly known as a forklift guidance system, or FGS). Operators or systems raise a material call-off. Drivers accept a claimable work order on a device. Acceptance, place, route, duration, and distance leave a record on the live hall instead of a radio memory.
Unnecessary patrolling rounds fell by about 95%. The plant recovered about 20 minutes of production time every day. Line micro-stoppages tied to late material were eliminated. Driver workload also evened out across the fleet as open jobs became claimable instead of radio-assigned. Recovered line time added about €5,080,000 (about $5,588,000) of production value in the first year. That figure is production uplift on this plant class. It is not a generic OPEX average and not a real-time location system (RTLS) fleet-cut number from a different module. More detail on the recovered minutes is in twenty minutes of production back every day.
The spend bought a real-time information layer on top of whatever MES, WMS, or ERP the plant already runs. Open requests and current mover locations sit in one view. Supervisors no longer need a physical patrol only to learn why a line is waiting. Product scope for claimable pull sits on material order automation.
Visibility is not autonomous control
Dark-factory pitches often blur two invoices. One invoice is spatial awareness and cleaner dispatch while people still sit in the seats. The other is replacing or fully guiding the vehicles themselves with an automated guided vehicle (AGV) or autonomous mobile robot (AMR) stack and closed-loop recovery. twinzo does not steer the trucks, does not replace PLC logic, and does not turn a human fleet into unattended movers. Treating those spends as one ROI number is how projects get oversold.
The AOS case shows what you get when the human logistics network becomes transparent and claimable. It does not show what you get when humans leave the network. That boundary matches what a true dark factory demands: observation and pull first, write-back authority later and only where the cell earns it.
Why recovered minutes still matter on a lights-out path
A true dark factory needs material to arrive without anyone searching for it. The intermediate step on a staffed hall is simply to stop the search. Twenty minutes of recovered line time every day compounds without the full automation stack. At the same time, closed calls and mover paths feed the spatially oriented dataset under an operational digital twin. Correlation and causality analytics can later ask which request gaps keep preceding a micro-stop. Live logistics recovery, stored place-and-time history, and pattern work are equal uses of the same record, not a ladder from search toward a predictive forecast pitch.
Plant teams can therefore review current traffic before they decide which corridors are candidates for fixed automation and which still need flexible human driving. The twin shows what is happening now. It does not simulate future physics or run discrete-event models. Those remain separate tools under simulation work. Path evidence for fleet size is covered in real travel paths on the floor and in internal logistics optimization.
How teams price the same bridge
1. Minutes drivers spend without an accepted task - Measure on a normal peak. Busy motion can still be empty search.
2. Count station stops caused by late material moves - Short stops still count even when no downtime code opens.
3. Price claimable pull against that baseline - Compare investment to recovered minutes and patrol share, not to an automation-scale AGV ROI.
4. Keep module stories separate - Do not average an AOS production-uplift case with an RTLS fleet-cut case. Scale examples that stay inside the visibility boundary are covered in why large plants still pay for empty rounds.
Commercial detail is on pricing and the features overview. For why most halls still run with people rather than full lights-out, see most plants are still full of people.
Get in touch if you want to walk through whether empty rounds on your hall are still an operating cost a claimable pull layer can remove before you spend on a lights-out cell.