Checking Vendor Nursing Hours Claims Before You Commit

What the BayCare hours claim actually says

On June 3, 2026, Relay Robotics issued a release about two robots added at BayCare's Winter Haven Hospital and Winter Haven Women's Hospital. The release says the robots are projected to complete 500 or more deliveries per month and carry an estimated savings of 150 or more clinical hours per month, collectively. Both phrases describe expectations, not a month that has already been measured.

That does not make the claim useless. It means a hospital leader should treat 150 hours as a hypothesis to test against their own runner logs, and should first find out what the vendor counted as an hour saved. The clearest way to do that is to collect a baseline before any lease quote is on the table.

The vendor's own figures do not line up neatly. An undated Relay case study page for BayCare lists 150 or more deliveries per day and 600 or more hours saved per month, across four robots at three hospitals. The June release covers two robots at two Winter Haven sites. Relay also says hospitals using its robots see average monthly time savings of over 150 hours, which is a company wide average and not a Winter Haven measurement.

Local coverage adds context. According to the Winter Haven Sun, the robots debuted May 6 and were funded by the Winter Haven Hospital Foundation through donor gifts. The Polk Region chief nursing officer said the aim was for teams without pneumatic tube access to spend more time in their departments. That is a workflow goal, and it is not an hours ledger.

Why a projection and a measurement diverge

Independent work gives a more sober picture. A 2026 single center observational study published in Scientific Reports, covering June 1 to November 30, 2025, found that robot delivery reduced delivery time by 32% to 36% compared with manual delivery. Routine tasks inside standard hours were included and emergency deliveries were excluded. The study reports that a 10 robot fleet replaced the original 19 full time logistics personnel.

The same study notes that excluded cases were less than 5% of total tasks. That matters because a vendor's hour count may quietly include or leave out the urgent work that pulls runners away from their routes, and that choice can shift the total more than the robots do.

Speed and support time cut both ways. A 2026 arXiv simulation, with a 30 day measurement period and a 200 bed reference hospital, used a robot speed of 54 m/min against a human speed of 72 m/min. Its two robot scenario increased overall late or incomplete requests by 0.37 percentage points. Human idle time fell by 11.5 minutes per 100 nominal beds per day, while robot support consumed 56.4 minutes on the same scale. The simulation is a model and not a hospital trial, but it shows the support work that a headline total can hide.

The economic evidence is thin. A systematic review of robots in critical care found that most studies lacked detailed economic analyses, and that some counted robot maintenance fees while others did not. If a study does not state its cost basis, its hours cannot be compared with yours.

Which baseline numbers to collect first

Collect these before any fixed payment lease is discussed. Four weeks of logs will show weekday and weekend patterns, and if your facility runs a night shift, it should be captured in full. Keep the log format identical before and after go live so the comparison is like for like.

• Trip count by route and by hour, split into routine requests and urgent requests.
• Minutes from request to pickup, and from pickup to handoff, including time spent waiting at the pharmacy, lab or nurse station.
• Route length, elevator rides and secured door passes for each trip.
• Late, failed or resent deliveries, the minutes each one cost, and who resolved it.
• Staff time spent on items that must stay with a person, so the share a robot could realistically take is clear.
• Time people spend supporting a robot once it runs, including loading, dispatch, clearing blockages and map corrections.

How long before the savings clock starts

Timing is easy to miss. The BayCare release, as carried by Yahoo Finance, says a typical hospital deployment, including integration hardware work, floor mapping and staff training, takes approximately 12 weeks, and onsite install time is 1 to 2 weeks. A baseline collected during planning is still valid, but a monthly savings figure that starts at go live should be checked against the date robots actually began routine work.

Check the delivery count as well. A figure of 500 deliveries per month should be split into trips the runner already made and trips that were never on the runner's list before. Only the first group is a labor saving. The second group is added capacity, and it deserves its own line in the business case.

How to label a figure before it reaches the budget memo

Sort every number into one of three bins: projected by the vendor, measured in a pilot by your own team, or measured by a third party. Only the third bin deserves a line item without caveats. A useful example comes from an outpatient pharmacy time and motion study in MDPI, with an external observer and before and after measurements. Daily staff time on stock intake, storage and order selection fell 59.3%, from 01:36:15 to 00:39:10.

That is a measured change in a defined set of tasks, which is exactly what a headline hours figure often lacks. It is also a reminder that a credible result is usually narrower than the marketing total. A clinic should expect its own pilot to produce a specific, bounded number, and should treat anything broader as a hypothesis until the logs say otherwise.

Where a lease quote enters the process

A lease quote should follow the baseline, not lead it. Service Robot Co. leases robots only, through financing on a three year lease, with one fixed payment quoted after a site assessment. The site assessment maps routes, door and elevator constraints and charging locations, so the payment and the projected hours are tested against the same floor plan.

We are OEM neutral. We pick the right robots across manufacturers, then finance, deploy, integrate, train and service every unit through a nationwide US engineer network. That means one partner handles the whole lifecycle, and the comparison of hours is not limited to a single manufacturer's projections.

Ask any vendor, including us, for the calculation behind its hours, the period it covers and the tasks it counts. A vendor that answers those questions plainly has given you something you can check against your own logs.