Scenario cookbook

Practical studies you can reproduce. Each recipe states the question, the setup, the steps and what to look at in the results. Most of them start from the 20 m loop built in Getting started.

A method that works for every study
  1. Save a baseline project, then change one input at a time.
  2. Use the same Run duration for every run (8 h is a good default; 24 h for batteries).
  3. Let the Run comparison table collect the runs and export it as CSV when you are done.
  4. Use Project›Save as to keep each variant you want to show.
  5. Every run is also kept in the Run archive: reopen it later and download a PDF report for your audience.

1. How many AGVs do I need?

BeginnerFleetRun comparison
Question
What is the smallest fleet that serves one order per minute without a growing queue?
Setup
The one-way loop from Getting started: one pickup, one drop-off, one charger and one parking. Add two more Charger or Parking stations on the bottom side, spaced a few metres apart, so up to four AGVs can be placed.
Steps
  1. Set AGV count to 1 and press Run (8 h).
  2. Repeat with 2, 3 and 4 AGVs.
  3. Open Run comparison in the dock.
Look at
Throughput, maximum queue, lead time P95 and utilisation.

How to read it. With one AGV the queue grows for the whole run and the lead time reaches hours: the fleet is undersized. From two AGVs throughput matches the demand (about 60 missions per hour), the queue stays near zero and extra vehicles mostly add idle time. The right size is the smallest fleet whose maximum queue stays bounded and whose P95 meets your target. Keep some headroom: a utilisation around 70–85 % absorbs peaks better than 100 %.

2. How much demand can the system handle?

BeginnerOrders
Question
At which order rate does the current fleet saturate?
Setup
The loop with two AGVs.
Steps
Run 8 h with Interval 60 s, then 50, 40 and 30 s.
Look at
The Queued orders chart in Results → Timeline, and maximum queue in the comparison.

While the system keeps up, the queue chart oscillates around a low value. Once demand exceeds capacity, the chart becomes a rising line and throughput flattens at the fleet's maximum: that value is the capacity of the layout with this fleet. Try the same test after raising Straight speed or reducing pickup and drop-off times to see which lever matters most.

3. Which dispatch rule works best?

IntermediateDispatchAllocation
Question
With several pickups and drop-offs, which combination of mission priority and AGV selection gives the shortest lead times?
Setup
A layout with at least three pickups and three drop-offs spread over the network, and three AGVs. Select all pickups and all drop-offs in the Scenario.
Steps
  1. Run with Oldest reachable mission + Nearest to pickup (the defaults).
  2. Change only the AGV selection to Fastest delivery, then Least busy time.
  3. Switch the priority to Pickup nearest to a free AGV and repeat.
Look at
Lead time average and P95, queue wait, and the loaded share of Fleet distance.

“Nearest” rules usually reduce empty travel and average lead time, but they can make distant orders wait longer, which shows in P95 and maximum queue wait. “Oldest” is fairer. Least busy time spreads the work evenly across vehicles; check the per-AGV bars in Fleet usage.

4. A dead end that blocks the fleet

IntermediateTrafficGridlock
Question
What happens when several AGVs share a two-way dead end?
Setup
Start a New scenario and draw four two-way 5 m paths forming a cross centred on (0, 0); a Pickup station at the centre, a Drop at the far end of the north arm (0, 5000), Parking locations at the ends of the west and east arms, a Charger at the end of the south arm. No links.
Steps
Set pickup = centre, drop-off = end of the north arm, Interval 1 s, Pickup 0 s, Drop-off 5 s. Run with 1, then 2 and 3 AGVs, watching the canvas at 5× or 10×.
Look at
The ‖ sign on waiting vehicles, Waiting for traffic in Fleet usage, the Waiting hotspots heatmap and the Possible gridlock alert.

With one AGV the system works (about 170 missions in the first hour, with a queue that grows because an order arrives every second). With two or three AGVs the fleet locks up within seconds: a vehicle leaving the arm meets one entering it, both wait, no mission is ever completed and Possible gridlock appears. Collision avoidance keeps the vehicles apart, but nobody organises who backs out first. More vehicles do not add throughput here: they add waiting. Fixes to try: make the arm exclusive with a traffic zone (next recipe), or give the drop-off its own one-way loop.

5. Protecting a single-lane corridor

IntermediateTraffic zone
Question
Does an exclusive zone on a narrow corridor remove the conflicts?
Setup
A T-shaped layout: start from recipe 4 and delete the south arm together with its Charger, so the Parking locations at the west and east ends are the two starting stations. The pickup is at the junction and the drop-off at the end of the 5 m north corridor. With Zone (9) draw a rectangle from about (−800, −800) to (800, 5500), covering the corridor and its entrance. Use 2 AGVs, an order every 12 s, 0 s pickup, 5 s drop-off and 100 cm/s straight speed.
Steps
Run 1 h with the zone. Then select the zone, delete it, and run again.
Look at
Completed missions, traffic waiting, the Entry waits heatmap, and whether a gridlock appears.

In our test, with the zone the two AGVs complete about 150 missions in one hour with no gridlock; without it they lock each other in the corridor almost immediately and complete a single mission. With the zone only one AGV enters the corridor; the other waits at the protected boundary, clear of the exit, and enters as soon as the first has left. The waiting moves from inside the corridor to its entrance, where it does no harm. When drawing your own zones, include the whole section where passing is impossible plus the entrances up to the point where vehicles can wait without blocking the exit.

6. Lead-acid or lithium?

IntermediateBattery
Question
How much capacity is lost to charging, and does opportunity charging with lithium help?
Setup
The loop with three AGVs and three Charger stations, an order every 40 s.
Steps
  1. Run 24 h with the battery model off (the reference).
  2. Enable Simulate charging and discharging, Lead-acid, 30 % → 80 %, 8 h autonomy, 8 h charge. Run 24 h.
  3. Switch to Lithium (2 h charge). Run 24 h.
  4. With lithium, try 40 % → 60 % to charge more often for shorter periods.
Look at
Throughput and P95 against the reference, the Charging share in each AGV's Fleet usage bar, and the queue timeline during charging periods.

Long lead-acid charges take whole vehicles out of service for hours: if throughput drops or the queue grows during those periods, you need an extra vehicle or more chargers. With shorter charges the fleet stays closer to the reference. Change one threshold at a time and remember that the curves are illustrative.

7. Two production lines with Plant logic

AdvancedPlant logic
Question
Can the fleet serve two lines with different rhythms, and which line suffers first?
Setup
Two pickup stations (line A and line B) and two drop-offs (warehouse A and B) on a loop, three AGVs.
Steps
  1. In Plant logic, build two independent sequences:
Periodic timer90 sGenerate missionA → warehouse AEnd sequence
Periodic timer150 s, first after 30 sGenerate missionB → warehouse BEnd sequence
  1. Name the mission blocks Line A and Line B, so generated orders say where they come from.
  2. Tick Use workflow to generate orders, go to Simulate and run 8 h.
  3. Shorten line A's timer to 60 s and run again.
Look at
Station rankings in Bottlenecks (queues per pickup station), lead time P95 and the queue chart.

In the comparison, runs driven by Plant logic show Diagram instead of an interval. The stations ranking tells which line's orders wait longest.

8. Replenishing a buffer with a counter

AdvancedPlant logicVariables
Question
A machine produces one pallet every 2 minutes into a buffer; an AGV should move one pallet at a time, only when one is available.
Setup
A pickup (the buffer) and a drop-off (the warehouse), two AGVs.
Steps

Production sequence:

Periodic timer120 sUpdate variablepallets +1End sequence

Transport sequence, looping:

StartPrerequisitepallets ≥ 1Update variablepallets −1Generate missionWait for deliveryWait5 sback to Prerequisite
Look at
In both mission blocks, select the buffer as pickup and the warehouse as drop-off. If a complete transport takes less than two minutes, throughput settles at 30 missions per hour. If the counter keeps growing, transport is the bottleneck.

To serve the buffer with two vehicles in parallel, add a second transport sequence identical to the first: both share the pallets variable. For the simpler case “serve the source whenever it is empty” a single Cyclic mission block is enough.

9. Chained transports: machine then warehouse

AdvancedPlant logic
Question
Material goes from the inbound bay to a machine; when it arrives at the machine, it must continue to the warehouse after processing.
Setup
Station 1 inbound (Pickup), station 2 machine (Pickup + Drop), station 3 warehouse (Drop).
Steps
Periodic timer120 sGenerate mission1 → 2End sequence
On station drop-offstation 2Wait60 s processingGenerate mission2 → 3End sequence
Look at
Lead times, and the queue at station 2 in the Bottlenecks ranking.

10. From a CAD drawing to a first run

BeginnerCAD
Goal
Model a real building from its DWG.
Steps
  1. Background›Import floor plan, check Applied units in the preview, then Use floor plan.
  2. Check the scale with Measure on a known dimension, such as a column spacing. Correct it with Scale factor if needed.
  3. Lower Opacity and hide unneeded layers with the Layers tool so the network stands out.
  4. Draw the aisles with Path and Ortho, add stations at docks, machines and chargers, and set directions.
  5. Configure the scenario and run.