Lens 03 · Proposal simulation

Demand in.
Fleet out.

Most AMR proposals hide the arithmetic. This one runs it in your browser, shows every constant, and prints a proposal you can take to a room. Change an input and the answer changes with it.

Runs locally — nothing sent Assumptions on screen Printable proposal

The job

20 m400 m
0.42.0

Runs entirely in this page. Nothing is submitted, stored or sent.

Simulated proposal

—

Estimate, not a quote
Recommended body

—

—

— AMRs required
— Cycle time per round trip
— Moves / robot / hour at planned speed
— Fleet utilisation target 85%
— Moves per day —
— Distance driven / day loaded and empty
— Install window survey to live
Utilisation —
Payload fit —
Shift vs runtime —

Deployment proposal — summary

01 Scope
—
02 Body
—
03 Fleet
—
04 Charging
—
05 Deployment
—
06 Next step
Book the floor survey. Every number above is an estimate until the route has been measured — the survey is what turns it into a commitment.
Have us check it
What the numbers assume

These are the constants and the formulas. Change any of them and the answer changes — which is why they are printed rather than hidden.

  1. Cycle = travel + handling. Travel is 2 × distance ÷ planned speed, because the robot returns empty on the same route.
  2. Handling is load time + unload time + 20 s of approach-and-align, taken as 120 s for pallets, 90 s for cartons and bins, 60 s for totes.
  3. Availability factor 0.92 — the cycle is divided by this to account for charging interruptions, faults and operator intervention.
  4. Fleet = ⌈ demand per hour ÷ (capacity per robot per hour × 0.85) ⌉. The 0.85 is a utilisation target, not a physical limit: a fleet run at 100% has no slack for a blocked lane.
  5. Planned speed defaults to 55% of the model's rated top speed, allowing for acceleration, docking and safety slowdowns. Fork models without a published top speed start at 1.2 m/s.
  6. Charging assumes rotation charging: one charger per two robots when runtime covers the shift, one per robot when it does not.
  7. Install window starts at 2 weeks and adds half a week for fleets above 4, above 10, and for goods-to-person station rebuilds.
  8. Not modelled: door and lift cycles, WMS handshake latency, traffic congestion between robots, and peak-hour clustering. All four can move a fleet size by one.
Before you trust this

A simulator is
not a survey.

This tool is honest about its inputs and honest about what it leaves out. What it cannot know: how wide your aisle actually is, how often a pallet arrives broken, whether the Wi-Fi reaches the far rack, and whether the shift change turns the main aisle into a crowd.

It is good for one thing: telling you whether the business case is in the region of two robots or twelve, before anyone spends money finding out.

  • Use your measured distances, not the ones in a drawing that predates the last re-layout.
  • Use peak demand, not average. A fleet that copes with the average fails at 6 a.m.
  • Round up, then check the exception path. The fleet size is the easy number; what happens when a station is blocked is the hard one.

Send the real numbers.

Take the estimate to your team, then send us the measured route. We will come back with a fleet size we can stand behind.

Request a Demo Back to the examples