Counting the Walking Hours a Delivery Robot Returns

What does the hospital study actually measure?

A six month study at a tertiary hospital in China gives the clearest recent number for turning delivery time into labor. The paper appeared in Scientific Reports in April 2026. It is a single center, parallel controlled study, and robot delivery reduced delivery time by 32% to 36% compared with manual delivery (p < 0.001).

The 10 robot fleet completed 7.3 times more delivery trips than 19 manual workers over the six months. The methods say the fleet replaced the original 19 full time logistics personnel, a ratio of about 1.9 employees per robot. The study details in this article come from the peer reviewed paper, not from the market report that first pointed us to it.

Why does the 7.3 figure need a unit check?

A market report summary of the same study describes the fleet as completing 7.3 more trips per week. That is a different claim from the paper's 7.3 times more trips over six months. A multiplier and a weekly addition answer different questions, and an operator who copies the wrong one will overstate or understate capacity.

Quote the paper's wording internally and keep the report as a pointer to the study. The same report also gives a market projection of USD 9.75 billion by 2034 at a 15.7% CAGR. That is a forecast from the report, not a verified primary figure, so it should stay out of your staffing math.

How do you turn minutes saved into staff hours?

Start with a baseline you can actually log. The unit of analysis is the delivery trip, not the shift, so the arithmetic begins with trips and ends with hours. The steps below keep the chain honest.

• Count completed manual trips per shift over a representative window, including return legs and urgent requests.
• Time each trip from request to confirmed handoff, and log the walking segment separately from elevator waits and pharmacy window queues.
• Apply the study's 32% to 36% reduction only as a starting assumption, then replace it with the figure from your own pilot.
• Multiply trips by minutes saved per trip, convert to hours, and count only the hours a supervisor confirms were redeployed to other work.

Does 19 to 10 mean nine jobs disappear?

The study credits the fleet's efficiency to uninterrupted operation, with no rest breaks, sick leave or shift rotations, and to path algorithms that cut idle trips. Those are reasons a fleet can outrun a crew, and they are not a payroll cut on their own. Hospitals usually redeploy freed hours to other tasks rather than eliminate budgeted positions.

The fleet ran SLAM based autonomous mobile robots, model R0033-A1, across 22 open wards. Human workflows also carry side effects that belong on the offset side of the ledger. A single site feasibility study at Korea University Guro Hospital, covering 122 non urgent missions in June 2025, found delivery success sensitive to elevator congestion. Nurses and pharmacists reported micro tasks such as walking farther to meet the robot, while support staff saw a substantial cut in manual carrying.

How should the cost comparison be built?

Compare two streams over the same period: the labor cost of the manual runs you replace, and the full cost of the robot system including service and integration. The study reports a 10 year lifecycle saving of 6.928 million RMB, with sensitivity analysis. That total reflects a Chinese hospital's setting and currency, so borrow its method rather than its number.

An older simulation by Rossetti, Felder and Kumar from 2000 modeled six robot couriers replacing three human couriers per shift. It showed a 34% decrease in turnaround time and break even at nine or more FTEs eliminated. It is a simulation, not a field result, so treat its break even as a hypothesis to test on your own floor.

For US facilities, Service Robot Co. leases robots on a three year lease with one fixed payment quoted after a site assessment. That turns the comparison into a fixed line set against measured hours, which keeps the math honest. We are OEM neutral, so we match robots from different manufacturers to the route, then deploy, integrate, train and service every unit through our nationwide US engineer network.

What belongs in the math besides hours?

Accuracy has a cost too. The robot group in the study reached 100% verification accuracy and item integrity, with a significantly lower omission rate (p = 0.011). Nurse and patient satisfaction were also significantly higher in the robot group (p < 0.05). Hours alone will not capture that, so log omissions and rework minutes during the same baseline window.

Elevator dependence matters as well. If your routes wait on lifts during morning medication rounds, robot trip times may stretch in that rush, and the baseline must cover that exact window rather than an average afternoon.