Hospital Delivery Robot Pilots: Planning for an Unknown End Date

What Duke's pilot tells hospital teams right now

Duke University Hospital in Durham, North Carolina, is running a pilot with three delivery robots that carry supplies through the building. Local outlets reported the launch on October 1, 2026, and WUNC followed on October 16 with a caveat that matters more than the launch itself: it is unclear how long the autonomous robots will be used at Duke Health.

For planners, the practical answer is this. Do not wait for a fixed end date before designing the pilot. Write down the conditions that would end, extend or expand it, set money checkpoints that finance can act on, and put review dates on the calendar before the first run.

The rest of this piece covers what Duke has said publicly, why open-ended hospital pilots are normal, and how to build exit criteria, budget checkpoints and review dates around one.

What Duke has said about its three robot fleet

CBS17 reported on October 1 that three delivery robots were transporting supplies throughout Duke University Hospital. Duke's team members submitted more than 280 ideas for the robot names, which are FloBot, R2-DU and WALL-D, according to a Duke Health news item reproduced by Find Health Clinics.

Operationally, staff must scan their badges on the robots to access items, and each robot has three secure compartments, according to WUNC. Hey Durham reports the robots can open doors, ride elevators and push floor buttons on their own. Each of those abilities is a dependency that a pilot has to test in the real building.

Chantal Howard, Duke's chief nursing and patient care services officer, described the purpose as taking trips that otherwise pull nurses, pharmacists and supply chain workers away from their work areas. Nurses can stay near patients instead of hunting for supplies, and pharmacy staff can send medications to the bedside, per Hey Durham. Howard also said the campus is massive, which means route length and trip frequency will decide how much value the fleet can return.

Why an open end date is normal for hospital pilots

A hospital usually launches a pilot before it can answer the questions that would set an end date. Those questions include how many trips a day a route really carries, how staff adopt the handoff, how the robot behaves through shift changes, and whether the time saved shows up on the ward.

Duke's own framing points the same way. Howard said the health system is expanding and redesigning models of care, which describes a program meant to evolve rather than a one-time test. An honest statement of uncertainty is more useful to a planner than a deadline nobody has measured against.

Clinical units also carry non-obvious obligations. Duke's official post says the robots make rounds to nursing units day and night, and that therapy dogs from Pets at Duke met the robots first so they would not be startled by its movement or sound. A pilot plan should list those obligations, not assume they will sort themselves out.

Setting exit criteria before the first run

Exit criteria are conditions written before go live, each with a measurement method and an owner. They work only when the team agrees on them in advance, because a pilot that is already running tends to justify whatever it is doing.

Three outcomes are worth writing down. Each one needs a threshold and a date for the decision.

• Stop: a safety event traced to robot movement, or a route that staff abandon after the early weeks despite training.
• Extend: trip data is still too thin to judge value, but no safety issue is open and the route is still in use.
• Expand: trip data shows the robots carry load that otherwise pulls clinical staff away, and the team can name the next units to add.

Budget checkpoints that finance can act on

A budget checkpoint is a date and a decision. It tells finance whether to continue, change or pause spending while changing course is still cheap. Good checkpoints sit after the first full week of routine runs, at the end of the first month of operation, and before any commitment beyond the initial term.

Tie each checkpoint to the operating numbers the pilot must prove, such as completed trips, staff time returned to patient care and failed runs. Do not tie it to a robot's specification sheet, which says little about your corridors, elevators or shift pattern.

Keep the cost side simple. A single known payment, set before the pilot starts, lets finance compare the robot against the staff time it is meant to free instead of chasing variable charges.

Review dates and governance that keep a pilot honest

Put review dates on the calendar with named attendees. Nursing, pharmacy, supply chain, facilities, IT for badge and network issues, and safety should all have a seat. Hold a short review at each checkpoint, a longer one at the decision date, and record every decision with its reason.

Include the people who meet the robot every day. Nurses and pharmacists see the failed runs and the workarounds first, and a review that hears only from the project sponsor will miss them.

Write the pause rule now, not after an incident. It should say who may take a unit out of service, how staff fall back to a manual courier, and who tells the rest of the building.

Where a service partner fits the pilot timeline

Service Robot Co. is a full-service commercial robot integrator for US businesses. We lease robots only, on a three year lease with one fixed payment quoted after a site assessment. The assessment is where exit criteria get their numbers: route length, door and elevator dependencies, and the layout data the robot needs to run.

We stay vendor neutral, so the robot is chosen for the building rather than for a brand. Our team handles deployment, integration, staff training and service through a nationwide US engineer network, so the pilot has one partner and one number from the first survey to the review that decides its future.