toolbox
Home / AI, Data & Technology / Delivery Robot Fleet Sizing Calculator

Delivery Robot Fleet Sizing Calculator

Estimate the delivery-robot fleet required to complete a route workload within a fixed service window. Instead of treating every scheduled hour as productive, the calculator applies a utilization factor to each robot’s delivery rate so the fleet estimate can reflect curb access delays, customer handoffs, waiting, repositioning, and charging interruptions.

The result is useful for last-mile pilots and campus or district delivery programs where managers need a fast capacity screen before route-level simulation. Because delivery demand is often uneven by hour and geography, the rounded fleet should be viewed as a baseline for the entered period. Weather, pedestrian traffic, elevator access, geofencing, dispatch rules, and service-level targets can all require more reserve capacity than the simple average model shows.

Workload assumptions

deliveries
deliveries/hr
hours
%
Result
—
Minimum delivery fleet
Robots required—
Effective tasks per robot—
Fleet effective capacity—
Spare effective capacity—

1. Enter required deliveries
Use the deliveries that must be completed inside the same service window you are sizing.

2. Set a sustained delivery rate
Enter completed deliveries per robot hour from a route or pilot that resembles the planned operating area.

3. Define usable operating hours
Use the hours each robot can actually be dispatched during the service window.

4. Discount for route inefficiency
Apply utilization for waiting, repositioning, access delays, charging, and other non-delivery time.

5. Check the whole-robot requirement
The calculator rounds upward so modeled capacity is not below the requested delivery count.

6. Stress-test peak periods
Run a second scenario for concentrated demand if the daily average does not represent the busiest hours.

Effective capacity per robot = Tasks per robot hour × Available hours × UtilizationMinimum fleet = ceil(Tasks required / Effective capacity per robot)Spare capacity = Fleet × Effective capacity per robot − Tasks required

Where:

  • Tasks required = workload that must be completed in the selected period
  • Utilization = share of available robot time that produces completed tasks
  • ceil = round upward to the next whole robot

Assumptions: All robots are modeled with the same average rate and availability. No explicit reserve is added for maintenance failures, peak-hour concentration, or charging-station constraints.

What the result means

The result is the smallest whole-number fleet whose modeled effective capacity meets or exceeds the entered workload.

For high service levels, test peak-period demand and consider a separate reserve or availability factor.

Given:

  • 520 deliveries per period
  • 3.2 deliveries per robot hour
  • 9 available hours
  • 68% utilization

Calculation:
Effective capacity per robot = 3.2 × 9 × 0.68 = 19.584 deliveries. Fleet = ceil(520 / 19.584) = ceil(26.55) = 27 robots.

Result:
Minimum operating fleet = 27 delivery robots; modeled effective capacity ≈ 528.8 deliveries.

Interpretation:
The baseline fleet covers the entered workload on average; weather, access delays, or charging constraints may justify additional reserve capacity.

Should I use deliveries requested or deliveries completed?

Use the demand the fleet is expected to serve, including the deliveries required by your service target. If cancellations are material, model them consistently in the demand forecast.

How should failed delivery attempts be handled?

If failed attempts consume meaningful robot time, reflect them in the sustained delivery rate or utilization. Do not ignore them if they are common in the operating area.

Does the estimate account for route distance?

Not directly. Route distance and stop spacing affect the deliveries-per-hour input, so use a rate measured or simulated for a comparable geography.

How do charging constraints affect fleet size?

Planned charging can reduce usable hours or utilization. If charging infrastructure creates queues, consider a lower utilization assumption or an explicit reserve fleet.

Why size to a peak window instead of a full day?

A fleet that can handle average daily demand may still miss service levels during lunch, evening, or other concentrated peaks. Matching the calculation window to the operational bottleneck gives a more useful result.