Building deliveries
Deliver 24 parcels from the mailroom to two office floors using one shared lift.
No sign-in needed: open it in Studio, inspect the saved results and edit in 2D and 3D. Sign in to connect your own AI; project files remain free. Check Account for execution access. Opening asks before replacing your current space.
How much sooner can 4 robots deliver the same 24 parcels than 2 robots with one shared lift?
- Who uses it
- Facilities teams and robot integrators
- How it can create value
- Compare waiting for a robot, waiting for the shared lift, final parcel handoff and the robots’ final return. Use a delivery-time target to screen fleet sizes before a trial, then check vendor costs and the real operating constraints.
- Bring your site data
- Parcel arrivals, loading and handoff times, floor routes, delivery target and vendor quotes. The synthetic lift holds one robot and uses fixed collective rules with travel and stop times. Passenger traffic, vendor dispatch, vehicle conflicts and charging are outside this comparison.
A model you can inspect.
Deliver 24 parcels from the ground-floor mailroom: 12 to each office floor, one parcel per floor arrives every 20 seconds. Compare 2 and 4 robots at 1 m/s using the same routes and one shared lift car. Each parcel waits for a robot, loads for 8 seconds, rides the lift and is handed over for 15 seconds. Delivery completes at handoff; the robot stays busy on its empty return. The lift holds one robot, moves at 0.8 m/s and spends 4 seconds operating doors plus 2 seconds dwelling at each stop, using the engine’s fixed collective rules. These are synthetic assumptions, not measured vendor dispatch. Job distance excludes empty returns; robot distance includes them. Vehicle conflicts, passenger traffic, charging and failures are not simulated.
Change one thing
Compare 2 and 4 robots: 12 parcels per office floor, one per floor every 20 seconds; the lift and routes stay the same.
Read the consequence
Separate parcels waiting for a robot from robots waiting for the lift. Delivery ends at handoff; the robot is ready for its next parcel only after returning. One lift car uses fixed collective rules, not measured vendor dispatch.
View the starting inputs
- Deliveries per office floor
- 12
- Order interval per floor (s)
- 20
- Parcel loading (s)
- 8
- Parcel handoff (s)
- 15
Agent counts, routes and resource definitions are also saved in the project JSON.
The same demand. A different capacity.
One run, seed 0. Observation window 60 minutes. Arrivals, geometry and the seed set stay the same in every option.
| Scenario | Completed | Robot wait per parcel | Lift wait per ride | All parcels delivered | All robots home |
|---|---|---|---|---|---|
| 2 delivery robots | 24 | 10m 00s | 8s | 25m 21s | 26m 16s |
| 4 delivery robots | 24 | 4m 09s | 13s | 13m 41s | 14m 37s |
Robot wait is averaged over 24 parcels; lift wait over 48 loaded and empty rides. Delivery finishes at handoff. Return time includes the robots’ empty trips back to the mailroom.
In this synthetic model, four robots deliver all 24 parcels 11m 40s earlier (25m 21s → 13m 41s) and cut the average wait for a robot by 58%. The shared lift is not the constraint yet: with four robots it is in use about 23% of the time, the average wait per ride rises from 8s to 13s, the longest wait is 51s and at most 3 robots queue. What remains is each parcel’s own trip: 8s to load, a lift ride of about 13s each way, 15s to hand over, and the routes at 1 m/s. Weigh the gain against the added fleet cost; passenger traffic, charging, robot-to-robot avoidance and dispatch rules are not modeled.
Waiting averages cover arrived jobs, including unfinished waits; cycle averages cover completed jobs only. These are synthetic calculations, not measured site performance, revenue or a safety assessment.
Engine, status and output notes
process-des-2.7.0. Status: 2 delivery robots: completed; 4 delivery robots: completed. “Completed” is the engine run status, not success for every job. Check blocked jobs and warnings.
- No engine warnings.
Engine inputs and saved comparisons stay with the saved project. Any limited trace affects playback details, not aggregate metrics.
Articles on this example
Plan in 2D. Inspect in 3D. Keep the whole project.
- Open and adjust. Switch between 2D and 3D; edit a part, a floor or an input. Both views describe the same model.
- Connect your agent for free. Sign in, choose Connect AI and connect ChatGPT, Claude, Codex or Claude Code. Ask it to inspect the example, edit the layout and check a route.
- Save a checkpoint. Project → Project file, or ask the agent to export the complete project. Import it later to restore floors, scenarios, saved comparisons and the selected view.
- Compare options. Run new operational comparisons under the same demand. Check Account for execution access, built-in AI and manual cloud-saving limits.