Saint Peter's Healthcare System in New Brunswick, New Jersey announced on July 1, 2026 that it has expanded its deployment of hellocare.ai's intelligent hospital room platform to cover the full organization — taking what began as a pilot deployment and scaling it to all patient rooms. The system is now running three concurrent modules: AI-assisted virtual nurse rounding and telehealth, AI-assisted virtual sitting, and active AI models for fall prevention that flag risk in real time.

Saint Peter's reports it is already seeing measurable impact on patient safety metrics, specifically in fall reduction and patient experience scores, though the health system has not released specific fall rate data. The platform is also providing clinical teams with real-time patient information visible from outside the room via intelligent digital door signs — a workflow tool that lets nurses and physicians see critical information without entering the room unnecessarily.

What the System Actually Does

The hellocare.ai platform combines several capabilities that have previously existed as separate tools:

  • AI-assisted virtual nurse rounding: A remote nurse or care team member conducts structured rounding via a two-way video interface in the room, reducing the number of physical rounding visits required while maintaining the clinical contact. The AI layer flags patients whose responses or behavior warrant escalation to in-person assessment.
  • Virtual sitting: Instead of placing a physical sitter in the room for patients who require continuous observation — a resource-intensive intervention — remote observers monitor multiple rooms simultaneously through the platform. The AI actively analyzes patient movement patterns and alerts the observer when escalation criteria are met.
  • Fall prevention AI models: Machine learning models analyze patient movement, position, and behavior in real time to generate risk scores and alerts. Unlike traditional fall prevention tools (Morse Fall Scale, call light response rates), this system attempts to identify fall risk before the patient initiates a fall sequence.
  • Intelligent digital door signs: Patient-facing information — isolation status, fall risk, diet restrictions, current care team — is displayed on smart signs outside the room that update in real time, reducing the communication overhead for handoffs and entries.

Context: Why AI Virtual Nursing Is Scaling

Saint Peter's is not the first or only health system deploying this type of platform. Ardent Health separately announced a partnership with hellocare.ai to deploy AI-assisted virtual nursing across 2,000 hospital rooms across its system, also in 2026. The acceleration is driven by three converging pressures: persistent nursing vacancies that make full physical staffing at traditional ratios unsustainable, rising costs of 1:1 sitter staffing, and the maturation of hospital-grade remote monitoring technology to the point where it can reliably handle the ambient observation task.

Virtual sitting specifically has seen rapid adoption because the math is straightforward: a 1:1 physical sitter typically costs $35,000–$55,000 per year in wages and benefits for continuous coverage of a single patient. A remote virtual sitter can observe 4–6 patients simultaneously. At scale, health systems are reporting significant cost savings from virtual sitting programs that fund further investment in bedside nurse compensation and recruitment.

For Bedside Nurses

AI virtual nursing is a displacement conversation worth having honestly. Virtual rounding and virtual sitting do reduce some of the task burden on bedside nurses — repetitive observation tasks, structured rounding documentation, routine check-ins. Whether that translates to improved staffing ratios (fewer patients per bedside nurse) or simply the same ratios with different task distribution depends entirely on how hospital administration deploys the savings. If your facility is adopting this technology, asking explicitly how the freed time translates to staffing decisions is a legitimate and appropriate question for your union, charge nurse, or nursing leadership.

The Clinical Evidence Question

The announcements coming from health systems deploying AI virtual nursing platforms are generally vendor-co-produced press releases describing positive outcomes in general terms. The clinical evidence base for AI-assisted virtual nursing is still in early stages: a 2026 study published in npj Digital Medicine found that virtual nursing discharge programs cut 30-day ED readmissions in one system, but the evidence for fall prevention AI specifically — the highest-stakes claim — is more mixed and methodologically variable.

The Saint Peter's announcement is notable because it describes a live full-system deployment rather than a pilot. Outcomes data that emerges from the full deployment — fall rates, staff satisfaction, cost per safe patient day — will be meaningfully more informative than the initial press release. For nurses evaluating whether AI tools in their facility are genuinely improving safety or primarily reducing headcount, asking for before-and-after fall rate data, sitter hours, and nurse staffing ratio changes is the right clinical question.

The broader trajectory is clear: AI-assisted virtual nursing is moving from innovation project to standard infrastructure. KLAS Research and other healthcare IT analysts project that within three years, the majority of acute care hospitals will have some form of continuous AI-augmented patient monitoring. The clinical and operational questions about how it's implemented — and who benefits from the efficiency gains — are the ones working nurses should be asking now, before the contracts are signed and the implementation is done.