Duke Health's Three-Robot Pilot: What Hospital Teams Should Measure

What did Duke Health actually report?

Duke University Hospital in Durham, North Carolina, is running three autonomous delivery robots, FloBot, R2-DU and WALL-D, in a pilot that moves medications and supplies around the building. WUNC's October 16, 2026 report is the fullest public account I found. CBS17 reported that the three robots were already transporting supplies through the hospital, with a timestamp of October 1, 2026. Heydurham also covered the rollout.

The public reporting does not say how long the pilot will run. WUNC states that it is unclear how long the autonomous robots will be used at Duke Health. That gap matters more than the headline. A three-robot test with no end date is an experiment, and a hospital team reading about it should treat it as one.

Why does nurse time sit at the center of the story?

Duke's August 9, 2026 Duke Heart Pulse newsletter said the robots were introduced through the Duke Care Innovation Collaborative and will carry medications and supplies throughout the hospital. The same newsletter said tasks like these can take up to 30 percent of nurses' time each shift. That number comes from the institution and has not been independently verified, so read it as Duke's own estimate.

Chantal Howard, Duke's chief nursing and patient care services officer, made the logic plain in that newsletter. Every minute a nurse spends walking across the hospital for supplies is a minute not spent with a patient. In WUNC coverage she described the campus as massive. Scale is the core argument for the pilot.

Frontline staff also had a hand in the project. A listing that syndicates Duke's own post says more than 280 Duke University Hospital team members submitted ideas for the robot names. It is a small detail, but it suggests that adoption tends to go better when the people walking the halls are involved early.

How do the robots work around staff and patients?

According to Duke officials as reported by WUNC, staff scan badges on the robots to access items, and each robot has three secure compartments. WUNC and CBS17 both report that the robots can open doors, ride elevators, push floor buttons and interact with people.

Duke's Working@Duke social post adds two operating details. The robots make rounds to nursing units day and night, and therapy dogs were introduced to them so the dogs would not be startled by movement or sound. A hospital robot shares its corridors with patients, visitors, animals and carts, so its behavior around living things belongs in the operating plan from day one.

What should an operations team measure during a test?

Start with a baseline, not the robot. Time every supply or medication run as it happens today, from request to bedside, and split the time into walking, elevator waits and handoff. Then take the same measurements during the pilot. Without that baseline, any claim about saved nurse time is only a guess.

Track at least these measures:

• Round-trip minutes per run, split into travel, elevator wait and bedside handoff
• Share of runs completed with no staff help at doors, elevators or floor buttons
• Accuracy of medication and supply delivery, including any compartment access errors
• Staff interruptions logged per shift, and whether the time spent walking moved to patient care
• Coverage by hour, since Duke's robots make rounds day and night

Where does this pilot fit the wider research?

A 2025 arXiv survey of healthcare delivery robots says they can deliver samples, meals, medications and medical supplies, and that they can support smoother operations and ease staff burdens. The same survey notes that robots carried medications and food during COVID-19, which minimized human contact.

A separate 2025 systematic review describes medical supply delivery as robots autonomously transporting medications, IV fluids, lab samples and surgical instruments. Those categories make a useful inventory for any hospital deciding where to start. Duke's public materials name medications and supplies, so a hospital should not assume its own pilot must cover every category in the review.

What belongs on a pre-pilot checklist for supply and medication runs?

Before anything rolls, settle the inputs. None of the decisions below appear in public reporting from Duke, so each one falls to the hospital team.

• Map every corridor, elevator and door the route touches, and mark the points where staff help is likely needed
• Decide which item classes are allowed in each compartment and which need a second check
• Assign badge access by role, and name who may open a compartment and when
• Name an owner for exceptions such as a stalled elevator, a blocked corridor or a dropped run
• Confirm elevator and door integration works with your building before go-live
• Brief every unit on how staff and patients should react when a robot approaches

How does Service Robot Co. approach a test like this?

Service Robot Co. is a full-service commercial robot integrator for US businesses. We are OEM-neutral, which means we pick the right robots across manufacturers, then finance, deploy, integrate, train and service every unit through a nationwide US engineer network. One vendor covers the whole lifecycle, so an operations team has one partner to call when a run stalls.

We lease robots only, through financing on a 3-year lease with one fixed payment quoted after a site assessment. For a hospital or any multi-floor facility, that assessment is where the mapping, elevator checks and measurement baseline described above get set up before the first run.