A two-bin kanban system is a visual replenishment method that holds a part in two containers: you draw from the first bin until it is empty, then start on the second while the empty bin itself becomes the reorder signal. The empty bin is the kanban, so there are no cards to lose.

Two-bin is the simplest pull system there is, and it is the right first pull system for most plants. It runs fasteners, gloves, labels, adhesives, and packaging without software, cards, or a planner. If you have read our broader guide to kanban in manufacturing think of two-bin as the entry-level, physical form of that idea: same pull logic, no cards to design or track. It is a small piece of lean manufacturing that pays off in the first week.

What Is a Two-Bin Kanban System?

A two-bin system stores one part number in two identical containers, each holding the same fixed quantity. Operators consume from the front bin. When it runs dry, they slide the back bin forward and keep working without interruption, and the empty front bin is moved to a marked return spot to trigger a refill. The refill has to arrive before the second bin empties. Because there are only ever two bins, the most inventory you can hold of that part is twice the bin quantity, which caps your on-hand stock automatically.

The elegance is that the container carries the signal. There is no card to fill out, no scan required, and no scheduler deciding when to reorder. When a bin is empty, you reorder; when it is full, you do not. That is the whole rule. This makes two-bin a natural piece of visual management: anyone walking the aisle can see what needs replenishing by looking at the empty-bin return rack.

The two-bin replenishment cycleThe two-bin replenishment cyclePOINT OF USEBIN Ain useBIN Breservedraw from A firstA emptiesEMPTY BIN= the signalreturn rack → supplier / storesREFILLmust beat bin BMax on-hand = 2 × bin quantity. The refill must land before bin B runs out.No cards, no scans, no planner: the empty container is the reorder trigger.
Consume A, slide B forward, and let the empty bin trigger the refill. The container is the kanban.

How Does the Two-Bin System Work Day to Day?

On the floor, the cycle is boring by design, which is what you want. An operator takes parts from bin A. When A is empty, they place it on a return rack or shelf marked for that part and start pulling from bin B. A material handler or supplier runner sweeps the return rack on a fixed route, refills the empty bins, and returns them to the reserve position. As long as the refill lands before bin B runs out, the operator never stops, never counts, and never fills out a form.

The one number you must protect is the timing: the empty bin has to be refilled within the replenishment lead time, and bin B has to hold enough to cover that lead time plus a safety margin. Get the bin size right and the system is self-correcting. Get it wrong and you either stock out (bins too small) or drown in inventory (bins too big). Sizing is the whole game.

Notice what the system quietly kills. It removes the counting, the min-max spreadsheet, and the weekly walk where someone eyeballs shelves and guesses what to order. It removes the panic buy when a line stops for want of a washer. And it removes the overstock that piles up when a nervous planner pads every order. Those are three different forms of waste, and two-bin attacks all of them with a piece of tape on the floor and a return rack. It is one of the cleanest examples of pull replacing push that you can put on a floor in an afternoon.

How Do You Size the Two Bins?

Sizing a two-bin system is a short calculation you can do on a napkin. Each bin must cover demand during the replenishment lead time, plus a cushion for variation. Work through these steps.

  1. Measure average demand. Pull the real usage rate for the part in consistent units per day or per hour. Use actual consumption history, not a guess or a standard that is three revisions old.
  2. Measure the replenishment lead time. Time the full loop honestly: how long from the moment a bin goes on the return rack until a full bin is back in the reserve position. Include the runner's route interval and the supplier's turnaround, not just transit.
  3. Compute the reorder point. Use N = (D × LT) + SS, where D is average demand, LT is lead time, and SS is safety stock. The reorder point is the amount you need on hand when a bin empties, which is why one bin's quantity should be at least this figure.
  4. Set the safety stock deliberately. A common approach is SS = (max daily usage × max lead time) − (average daily usage × average lead time), which sizes the cushion to the variability you actually see rather than a round number. Bigger swings in demand or lead time mean a bigger cushion.
  5. Round to a sensible pack. Set the bin quantity to the reorder point plus safety stock, then round up to a standard pack, box, or bag so refills are clean. Both bins hold this same quantity.
  6. Recheck when the loop changes. Re-size whenever demand, supplier lead time, or the runner route changes. A bin sized for last year's volume silently starves or floods the line.
Sizing one bin: demand, lead time, and safety stockSizing one binbin qty = (demand × lead time) + safety stockbin Abin Bbin Asafety-stock floorfull binEach empty bin must be refilled before the other one crosses the safety-stock floor.
One bin has to cover demand across the whole replenishment loop, plus a cushion for the days that run hot.

Which Parts Are Two-Bin Best For?

Two-bin shines on small, cheap, standardized, high-volume parts with steady demand: fasteners, o-rings, gloves, labels, tape, adhesive cartridges, shipping supplies. These are the parts nobody wants to plan and everybody stocks out on at the worst moment. Two-bin removes them from the planning conversation entirely, which frees your planners to spend their attention on the parts that actually deserve it.

Two-bin loops also tend to be the first pull signals you place after a mapping exercise: when a value stream map flags a supermarket of consumables feeding a line, a two-bin loop is often the cheapest way to control it. It is a poor fit for the opposite profile: expensive parts where twice the bin quantity ties up real money, parts with lumpy or seasonal demand that break a fixed bin size, parts with short shelf life, or slow-movers where a bin might sit for months. For expensive or critical spares, a counted approach or a card system with tighter control fits better, which is a theme we cover in spare parts inventory management. Match the tool to the part.

Good fit for two-binPoor fit for two-bin
Small, low-cost partsExpensive parts (2× ties up cash)
Steady, predictable demandLumpy or seasonal demand
Standardized, high-volumeSlow-movers, one-offs
Long shelf lifeShort shelf life / perishable
Cheap to hold a little extraCritical spares needing tight count

How Is Two-Bin Different From a Card Kanban?

The pull logic is identical; the signal is not. In a card kanban, a card detaches when a container is consumed and travels back as the reorder trigger, which lets you run more than two containers, split lot sizes, and route signals across a plant. In two-bin, the empty container is the card, which is simpler but limits you to a two-container loop at a single point of use. If you need finer control, multiple loops, or electronic signals, you graduate to cards and the sizing math in our kanban card calculation guide. Start with two-bin, move to cards only when the part demands it. A practical rule: if a plain physical trigger will do the job, do not build a card system to do it. Complexity in a replenishment loop is itself a form of waste, because every card, scan, and screen is one more thing to maintain and get wrong.

The math only holds if the demand and lead-time numbers you fed it are real. Bins sized from a stale standard drift out of tune quietly, and the first symptom is a line-stopping stockout on a two-cent part. Plants that track consumption and replenishment timing live can see a loop starting to run hot before it stops the line, and can re-size on data instead of after the outage. That visibility over the systems you already run, with no rip-and-replace is where a simple physical system like two-bin gets an early-warning upgrade. See how one plant made its material flow legible in the CLS case study.