care.ai

One core module depends entirely on the models. care.ai, part of Stryker since 2024 and sold within its Smart Care business and SmartHospital Platform, has three parts. Ambient Monitoring uses AI driven room sensors that apply computer vision to detect presence, motion and behavior, infer patterns, flag deviations and notify care providers, for uses such as fall prevention, turning adherence, bed occupancy and rounding compliance; without the models it does not exist. Virtual Nursing and Virtual Monitoring put remote nurses and sitters on video into the room from a virtual command center, which works without a model.

The system detects and notifies; people decide. The Ambient Monitoring page says the sensors autonomously detect presence, motion and behavior and notify appropriate care providers when a situation needs attention, and that the virtual command center automates workflows, monitors protocol compliance and orchestrates tasks and notifications, with virtual nurses supporting a curated task list or monitoring several patients at once. Henry Ford Health's pilot used virtual nurses for admission, rounding and discharge tasks without reducing bedside staff.

The thresholds are missing: nothing says when the system raises or withholds an alert, how many rooms one observer covers or what happens when two rooms need attention together, and nothing describes what follows a missed event.

The approach is described in general terms and nothing is identified. The Ambient Monitoring page says the sensors use computer vision and AI driven algorithms to detect presence, motion and behavior, infer patterns and identify deviations, and that the platform can predict issues before they escalate, with uses listed from fall prevention to bed occupancy. No model, training data, version, update practice, detection accuracy or false alarm rate is published.

No party behind the sensors or the cloud suite is named. The Smart Care and Ambient Monitoring pages describe AI driven sensors in each room, a cloud based suite, and a virtual command center connected to record, nurse call, documentation, interpretation and device systems; Henry Ford Health's deployment added SONIFI Health's in room platform. No model provider, cloud host or subprocessor list is named for care.ai.

Virtual nursing brings a remote person into the room on video, a different data path from sensor inference, and nothing published says where that stream goes, who can join it or whether it is recorded.

A named customer and no published results. Stryker's October 2025 news item says Henry Ford Health (13 locations, 50,000 staff) is expanding its virtual care model enterprise wide after a virtual nursing pilot with care.ai, in which virtual nurses handled admission, rounding and discharge tasks without reducing bedside staff, and lists its goals (cost savings, safety, outcomes, HCAHPS scores); the case study with any figures is released on request. The Ambient Monitoring page cites a 2023 McKinsey study estimating 15 percent net time savings from task delegation and technology, which is about the approach, not this product.

No outcome figure, detection accuracy or study of care.ai is published.

General assurances are published and the questions continuous room sensing raises are not answered. The Ambient Monitoring page says the sensors detect presence, motion and behavior "without intruding on patient privacy", and Stryker's privacy statement, which covers people using its products and services, says it maintains physical, technical and administrative safeguards and may share product and service statistics in de identified, aggregate form. The Smart Care page describes care.ai as a cloud based suite.

Continuous sensing captures patients who are confused, sedated or undressed, their families and staff. Nothing published states whether video is stored or processed transiently, how long anything is kept, what patients are told, whether they can decline, or whether captured material trains the models. Nothing is published on safety engineering.

A privacy document that reaches the products is published, and no business associate position is stated. Stryker's privacy statement applies to personal information collected when people use or buy its products and services, lists health information among the sensitive data it may hold, and describes physical, technical and administrative safeguards.

It takes no position on business associate status for patient data that care.ai handles for hospitals, and no care.ai specific privacy or agreement page is published. The acquisition also changes who a hospital contracts with.

Stryker publishes product security advisories and names no certification for care.ai. Its security advisories page lists vulnerability notices for its products, including Vocera servers, which shows a disclosure practice; the privacy statement describes physical, technical and administrative safeguards. Stryker also publishes a product security program page and a Vocera security and privacy overview, which may describe independent testing or attestations; the Vocera document covers a separate product line.

No SOC 2 report, ISO 27001 certificate, HITRUST certification or penetration testing statement is published for care.ai itself.

Ambient monitoring and virtual care are described as monitoring and workflow support, with no device claim: fall prevention monitoring, turning adherence, bed occupancy, rounding compliance and notifying staff, with virtual nurses and sitters watching the room. The owner is a medical device manufacturer and the platform sits beside regulated equipment such as beds and physiological monitors, and Stryker's alarm routing middleware is a separate product. Every alert arrives with a live view a person can check.

The ambient detection carries no published evaluation, monitoring or testing across patients. The Ambient Monitoring page says the sensors detect presence, motion and behavior "without intruding on patient privacy" and can predict issues before they escalate, and gives no accuracy, validation method or results for any use.

The exposure is specific. Vision based detection can vary with body size, skin tone, lighting, bedding and position, and the patients placed under continuous observation are often older, frail, confused or attached to lines. A model that performs worse for some of them concentrates its misses in the group chosen for observation because it was judged highest risk, and where virtual monitoring replaces a bedside sitter the alternative observation has already been removed.

Alerts reach people who can see the room, and nothing stands behind the output. The Ambient Monitoring page says the sensors notify care providers when a situation needs attention and feed a virtual command center where virtual nurses and sitters monitor patients, and that Virtual Nursing and Virtual Monitoring can run in the same room on the same sensor, so a person can check an alert against the live view before acting. That is a review mechanism.

No detection accuracy, false alarm rate, warranty, indemnity or correction route is published. For a system that decides which room a remote observer looks at, the missed event rate decides whether it helps, and the false alarm rate decides whether staff keep trusting it.

Integration is claimed for care.ai with no system named. The Ambient Monitoring page says the virtual command center can integrate with the EHR, nurse call, documentation tools, interpretation services and IoT devices, and that Stryker's implementation teams connect care.ai to existing systems; at Henry Ford Health it was combined with SONIFI Health's in room patient platform. No record system, interface standard or listing is named for care.ai. The broad integration library Stryker describes belongs to its Vocera communication and Engage middleware products, a separate product line on the same platform.

The delivery model is stated and location is not. Stryker's Smart Care and Ambient Monitoring pages describe care.ai as an integrated, cloud based suite, with AI driven sensors installed in each room and a virtual command center drawing on near real time data from them. Earlier company material described inference on the sensor unit itself; the current pages describe AI driven devices in the room and a cloud based suite without saying which processing happens where.

No cloud provider, region, residency commitment or statement of what leaves the building is published.

No price, unit of charge or pricing basis is published, and buyers are asked to contact an expert. The cost has hardware in it, since a sensor goes into every monitored room, and the product is one part of a platform designed to be adopted with Stryker's beds, communication badges and alarm middleware.

Coverage is named for the hospital inpatient unit, with one named deployment behind part of it. Stryker sells care.ai for inpatient rooms: virtual nursing for admission, rounding and discharge work, virtual sitting, and ambient monitoring for fall prevention, turning adherence, rounding compliance, staff duress and bed occupancy tracking, which Stryker sells to support patient flow and throughput. Henry Ford Health's pilot and enterprise expansion support the virtual nursing part.

Nothing published validates the ambient detections, no specialty breakdown is given, and nothing addresses settings outside the inpatient unit.