Rochester's Moxi Milestone: Which Figures Are Estimates

What did Rochester Regional Health report?

On August 21, 2026, Rochester Regional Health reported that its Moxi robots had passed 100,000 trips across Rochester General Hospital and Unity Hospital, according to Finger Lakes 1. Fingers Lakes Daily News carried the same milestone as 100,000 hospital deliveries. The system's breakdown shows more than 75,000 deliveries at Rochester General and more than 25,000 at Unity since 2023.

The milestone is a count the health system announced, and the coverage we reviewed did not include an independent audit of it. The staff hours figure is weaker still. More than 50,000 hours returned to the hospital is an estimate, and the coverage uses the words trips and deliveries interchangeably without defining either. Read the total as evidence of sustained use, and treat the hours as a claim to test before anyone quotes it.

Which figures are counts and which are estimates?

Rochester's announcement mixes two kinds of numbers. The trip and delivery totals describe completed work. The hours and walking figures are conversions, and a conversion depends on assumptions a reader cannot see from a press story.

The walking estimate is a good example. The system put employee walking eliminated at about 55,000 miles, described as about 120 million steps. Those figures estimate avoided movement rather than log it, and the coverage did not describe how the distance was derived. Treat the two numbers as one estimate expressed two ways.

What do the robots carry in these hospitals?

The system named three task groups: medications, supplies and laboratory specimens. Casey Wilbert, identified in the coverage as vice president of pharmacy and imaging services, said the robots take on tasks that do not require a person. That framing points to the work a medication transport robot handles well, which is predictable, repeated moves along known routes rather than bedside care.

The older record fills in the timeline. WXXI reported in October 2023 that Rochester Regional was training staff to work with new robot employees. Older coverage from WXXI and WHEC says the system began using Moxi in 2023 and operates eight devices. The coverage reviewed did not confirm that device count for the present date, so the eight devices should be read as a 2023 detail.

How is the staff hours claim built, and what is missing?

The more than 50,000 staff hours returned is the most quoted number and the least documented. The coverage gave no methodology for turning deliveries into hours, and no independent audit was found. A figure like this is usually built by assigning a time value to each completed task and multiplying it by volume. That method is reasonable as a starting point, but the time value carries nearly all of the result.

The period is missing too. The coverage did not say over which months or years the 50,000 hours accumulated. Without a period, the hours cannot be turned into an annual pace, and they cannot be set beside a shift schedule or a staffing budget. A reader who sees the figure in a board packet should ask for the period before the number goes any further.

How does the 2023 projection compare with the 2026 result?

Before rollout, WHEC reported that Wilbert's team had identified about 200 hours a week of work Moxi would be able to do. That was a projection, set before the robots ran a single route. It describes capacity, meaning the work that could be offloaded, not a measured return. The 2026 figure describes an estimated result after the fleet ran. The two numbers measure different things, and neither one verifies the other.

WHEC also quoted Wilbert saying Moxi will not replace people. Taken together, the projection and that statement point to redeployed time rather than eliminated roles. The distinction matters for any budget, because redeployed hours only become savings when a manager decides where they go.

What does the wider Moxi record show?

Moxi is made by Diligent Robotics. The Robot Report reported a milestone of 100,000 autonomous hospital elevator rides, and Diligent's November 2024 release counted more than 100,000 autonomous elevator rides across U.S. health systems. The Robot Report also reported 300,000 pharmacy deliveries with Moxi, a figure the company itself reported. CBS News, as of early 2026, said Moxi was at 25 hospitals nationwide.

These are vendor and press figures. They show that the platform has operated across many health systems, which matters to a team checking whether a vendor has real scale. They do not verify Rochester's hour returns, and they do not predict what another building will see on different elevators and corridors.

What should a team ask before using Rochester as a benchmark?

Use the Rochester figures as a set of questions for vendors and peer hospitals, not as a target to hit. Put these questions to the health system, or to any vendor quoting similar results.

• Define one trip or delivery, and say whether a multi stop route counts as one task or several.
• Name the baseline time per task used to convert deliveries into staff hours, and the source of that time value.
• State the period covered by the hours figure, and whether the total came from task logs, time studies or staff surveys.
• List the task groups in scope, and say which work was excluded from the count.
• Confirm how many devices ran during the period, given older reports of eight devices.
• Show what happened to staff workload on the same wing during the same months.

Where does a site assessment fit?

Service Robot Co. leases robots on a three year lease, and the team quotes one fixed lease amount after a site assessment. That assessment maps routes, elevator access, staffing gaps and task volume before any number goes into a budget. It is the step that lets a facility build its baseline from its logs rather than borrow one from a press release.

We integrate, train and service every unit through a nationwide US engineer network, and we work across manufacturers rather than one catalog. For a hospital weighing a hospital delivery robot lease, the practical next step is a short trial that logs each trip, times a shift, and tests the staff hours claim against measured data. The log, not the headline, should decide the case.