Key takeaways
- A robot's return is walking and waiting time you can measure, minus the staff time the robot adds.
- Rochester's 50,000-hour figure is a system estimate from one report, not an audited result.
- Log trips for a full week, including nights and weekends, before you estimate anything.
- Exclude trips the robot cannot take, plus the loading, handoff and exception time it still needs.
- Net hours are capacity, so decide in advance where those minutes will go.
The short answer for a unit manager
A hospital unit can estimate the nurse walking hours a delivery robot might return by logging its supply and specimen trips, timing a sample of them, and subtracting the staff time the robot still needs. The answer is a range you can defend, not a single headline figure.
Large claims exist. Rochester Regional Health reports more than 50,000 staff hours returned after 100,000 robot trips, but those are the system's estimates. Your unit's number comes from your own trip log, and it should be easy to check line by line.
What the Rochester figure does and does not show
According to a FingerLakes1 report dated August 21, 2026, Moxi robots at Rochester General and Unity hospitals passed 100,000 trips, and the system estimates they returned more than 50,000 staff hours. The same report puts the steps spared at an estimated 120 million. The figures are system-reported and not audited, and we found them in one report only.
The robots carry medications, supplies and lab specimens, so the trip mix overlaps much of what a nurse manager already tracks. What the report does not show is the method: which trips counted, what time per trip was assumed, and whether staff time for loading and handoffs was netted out.
Other hospital claims measure different things
Cedars-Sinai reportedly saved about 300 miles of nurse walking in the first six weeks after a rollout. It cited an estimate that nurses can spend 30 percent of their time fetching and gathering supplies. That is the hospital's estimate, presented in a newsroom story rather than a time study.
Endeavor Health Edward Hospital, per a local news report, saw robots travel about 48,000 miles and make about 103,000 deliveries, with a nursing director estimating more than 100 million steps saved. Diligent Robotics announced more than 100,000 autonomous elevator rides across U.S. health systems in November 2024. That counts rides, not staff hours, so it cannot stand in for an hours estimate. In the sources we reviewed, we did not find an independent time-motion study with a verified hours total.
Log trips before you time anything
Start with a record of what leaves the unit, not with a survey of how staff feel about walking. A trip log captures the runs a robot might one day take, and the same log shows which ones are worth moving first. Cover a full week, including a weekend and at least one night shift, so the pattern is not built from a single busy day.
Record each trip with the same fields so the results can be sorted later.
- Trip type: supply, specimen, medication or other
- Origin and destination, including floor and elevator use
- Start and finish times, recorded by the person making the run
- Whether the run was bundled with another task
- Who signed for the item at the other end
Turning trips into hours
The arithmetic is simple. Multiply the number of trips in each category by the average minutes per trip, then divide by 60. Keep walking and waiting separate. A robot can shorten the walk, while the wait at a pharmacy window or for a lab courier stays the same, and that wait may be the larger share of the time.
Time a sample of runs with a stopwatch or a phone timer, covering each trip type and each shift. Average the times by category rather than by the worst run you remember. A sample taken on a bad day overstates the return, and a sample taken on a quiet day understates it.
What to exclude so the estimate holds
Exclude trips the robot cannot take. Controlled substances that need two-person verification, oversized equipment, and items that cannot ride an elevator at the hour they are needed all stay with staff. Count only the trips a robot could realistically complete on your route.
Exclude time the robot does not remove. Someone still loads it, receives it at the far end, responds when it stops in a corridor, and clears exceptions such as a blocked door. A run that already sits on the path to a patient room may save little, so do not count it as a full trip. Net the robot's staff touchpoints against the walking it returns.
How the net hours get used
The net figure is capacity, and capacity only helps if the unit decides in advance where it goes. Rochester's report describes its hours as freed time. A nurse manager needs a plan that moves those minutes to bedside work, not one that quietly disappears into the next shift's backlog.
Service Robot Co. is an OEM-neutral integrator. We pick the right robots across manufacturers, assess the site, finance the robot on a 3-year lease, deploy and train, and service every unit through a nationwide US engineer network. Once your trip log is in hand, the site assessment turns it into one fixed lease payment quoted after the survey, which is easier to compare against the hours you measured than any vendor headline.
Frequently asked questions
Want a robot working for you?
Tell us the job and the site. We will recommend the robot, quote the lease, and keep it serviced.
Sources
- FingerLakes1: Rochester Regional Health robots reach 100,000 hospital deliveries
- Newswise: Robots help nurses get the job done with smiles and beeps
- Cedars-Sinai newsroom: Robots help nurses get the job done
- NCTV17 report on Endeavor Health Edward Hospital robots
- Diligent Robotics: Moxi reaches 100,000 autonomous elevator rides
