AI and Robotics Will Make Today’s Hospitals Look Archaic

Tom Polen argued that the next decade will transform health care, especially hospitals, quickly enough that current practice will soon feel outdated. His clearest example was intensive-care monitoring. BD, he said, is already using AI to predict what may happen to an ICU patient 15 or 20 minutes in the future, allowing staff to intervene before the event occurs.

For now, Polen said, the system can tell a nurse to intervene. The future he described is more autonomous: devices connected to AI intervening directly to keep patients well. He extended the idea beyond clinical intervention to hospital operations, imagining humanoid robots handling comfort and support tasks, such as bringing a pillow, where a trained clinician is not required.

The limiting factor is not whether the technology can advance quickly. Polen identified safe testing and implementation as the bottleneck. Health care, he said, is an area where “you can’t have errors,” even if AI may often deliver higher quality than human performance. The path he described is prudence: test the technology, then implement it where it is safe enough to do so.

That makes BD’s view of AI adoption narrower than the broad market excitement around “physical AI” and robotics. In Polen’s framing, AI enters clinical settings through uses that can be validated in an environment where errors have consequences: first prediction, then guided intervention, and eventually more autonomous devices.

The first hospital robotics use cases Tom Polen emphasized were logistical, not diagnostic. BD has partnerships with robotics companies that deliver medications up to hospital floors so nurses do not have to walk down to the pharmacy and bring them back.

Polen described the reception as positive because the automation removes menial tasks from highly trained people. In his framing, the first phase of robotics in hospitals is not substitution for clinical expertise. It is removing work that does not require a trained clinician.

That same logic applied to BD’s own workforce. Polen said the company is “probably not hiring fewer people necessarily,” but is able to keep staffing relatively flat while continuing to grow. BD is adding capabilities that medical-device companies historically did not employ in large numbers, especially around AI.

The company is applying AI across manufacturing centers, products, and commercial organizations. Polen described momentum inside the organization as employees see the impact and bring more AI talent into their teams. The efficiency story, as he told it, is less about immediate headcount reduction than about scaling the company with a flatter operating base and different capabilities.

Tom Polen described BD’s strategy around three pillars: AI-connected care, enabling care to shift into the home, and improving outcomes for chronic disease. Caroline Hyde framed the shift partly around demand pressure and insufficient supply, asking whether more health care will need to happen at home. Polen said there is “obviously a consumer interest” in moving care from the hospital into the home, and that this is one of BD’s strategic priorities.

In BD’s case, the home-care strategy was concrete rather than a broad slogan. Polen pointed to infusions at home, urinary incontinence, and oncology treatments at home. He also said a “digital pathway” between patients at home and hospitals will become ubiquitous, with data flowing back into institutional care.

The same home-care direction applies to medication access. Polen said BD robots are running “a lot of the retail pharmacies today.” If someone goes online and orders drugs that arrive at home two hours later, he said, there are not pharmacists in those large e-retailers manually counting pills into amber vials; BD robots are doing that work. His point was that automation is enabling a different way for people to access medications, in the home rather than through a store visit.

BD’s scale makes supply-chain resilience more than an efficiency problem. Hyde framed the company as a manufacturer of 35 billion medical devices a year, or roughly 400,000 a minute. Tom Polen said hospitals in 190 countries rely on BD products to care for patients.

Polen said the company has spent decades building redundancy into the system: secondary suppliers for raw materials, infrastructure attached directly to plants, railroad spurs, and even resin storage capacity prepared for hurricane season. The point was not merely that BD has a large supply chain. Polen described resilience as part of health care infrastructure itself. If a hurricane disrupts a resin factory, the company’s resin reserves are meant to keep medical-device supply flowing to hospitals.

In his account, that kind of preparation has moved higher on the agenda for hospital leaders. “There’s never been a bigger focus by health care CEOs, hospital CEOs, than there is today” on resilience, Polen said. The recent context Hyde raised — including the Strait of Hormuz and US-China relationships — sat behind the question, but Polen’s answer emphasized long-built operational buffers rather than a specific geopolitical forecast.

BD’s operating model, as Tom Polen described it, is rooted in Lean and Kaizen practices under the company’s “BD Excellence” label. Hyde connected that to operational excellence and partnerships, including one she referenced with Amazon. Polen’s answer focused less on Amazon specifically and more on how manufacturing discipline can transfer into health systems.

He said BD works with hospital partners on operational improvement, and that one large US hospital system had recently been invited to visit BD manufacturing plants to see how the company applies lean practices. In Polen’s view, the methods needed to run a large, sophisticated manufacturing network are relevant to hospital problems: quality outcomes, cost outcomes, and process improvement.

Hospitals are still relatively early in applying some of those approaches, he said, leaving “tremendous opportunity for improvement.” That expanded BD’s partnership thesis beyond devices or automation: the company sees a role helping hospitals import operating practices from high-volume medical-device manufacturing.

The same practical lens shaped Polen’s answer on global health systems. He pointed to best practices in Northern Europe and Asia, as well as strong health systems in the United States, but resisted naming a single model. One system may be especially good at oncology treatment; another may have outpatient or hospital-at-home care right. Polen said he does not think anyone has mastered the “entire end-to-end ecosystem,” because health care is too complex.

What has changed, in his view, is the speed at which systems can learn. AI, he said, creates a moment in which organizations can “get better faster” than before, and he sees health systems moving up the curve more quickly than at any other point in his career.