Vayyar Care

A learned model drives the analytics module, and the fall detection at the core is not described. Vayyar Care turns the reflections a 4D imaging radar receives into events a care team acts on: falls, room presence, time at rest and bathroom visits. The Vayyar Care FAQ says activity data gathered by the sensors is "analyzed by AI, and presented on the INsights platform", with real time fall alert verification. The company's May 2024 post on AI fall prevention equates its AI with machine learning, "all about pattern recognition", and says its fall prevention system is trained on fall history, activity monitoring and resident record data (history of illness, medications, age, gender, ethnicity) to generate fall risk scores. That is a learned component, and it runs the analytics and risk scoring module.

How the radar classifies a fall in the first place is not described on the Vayyar Care page, the FAQ, the privacy policy or the June 2020 launch post: fall detection from a point cloud can be built on trained classifiers or on fixed rules over height and motion, and the company does not say which. The detection alerts, delivered into nurse call systems, are what facilities buy first, and they may stand without the learned risk model.

The model is the engine of the analytics and fall risk module, with the detection method unstated.

Detections are delivered into systems a person already watches, and the rules that decide an alert are not published. On the facility route the company says detection data "gets delivered into the central nurse call system", with TekTone, Austco and other nurse call and care platforms as partners, so a fall alert reaches staff where escalation and response already run. On the consumer route the privacy policy (updated 28 July 2025) states that the product "does not send fall alerts to End Users and/or to emergency services, only to Amazon's Alexa App, Alexa Device and the Alexa Together Services; which are solely responsible for sending such alerts", which puts alert delivery with a third party.

Two parts of the oversight structure are unpublished: the criterion that turns movement into a fall alert (the operating point, behavior when the system is uncertain, and how more than one person in a room is handled), and any check on falls the system misses. The Vayyar Care page's "Receive immediate alerts when a resident experiences a fall" and "Monitoring every movement" describe presence rather than correctness, so they set the scope the oversight is judged against rather than making an absolute claim; the June 2020 launch post's "100% automatic" describes the absence of a button, not accuracy. The DCB0129 clinical risk management standard achieved in January 2026 requires a named clinical safety officer and a hazard log, and the hazard log is not public.

The sensing approach is named precisely and the model layer is not. Vayyar Care is a 4D imaging radar that builds a point cloud of the room from a wall or ceiling mounted sensor, and the company publishes the silicon behind it: its own radar on chip platform, specified as covering imaging and radar bands from 3 to 81 gigahertz with up to 72 transceivers and an integrated digital signal processor. The claims can be tested against those limits. Coverage is published as a number, 13 feet to the front and 6 feet 6 inches to either side, with a statement that large rooms may need more than one device, and the company states that no optical data is collected.

What produces the classification is not identified. No model, architecture, training data, detection accuracy, false alarm rate or operating point is published for falls, presence, rest or bathroom visits, and nothing describes how or when detection software on installed sensors is updated. The June 2020 launch post's "4X more accuracy than other automatic fall alert systems" names no comparison, method or source.

The radar and the company's own silicon are named in full, while the model's identity and its versions go unnamed.

The chain is structurally short at the sensing and classification layers and open at the data layer. Vayyar designs the radar on chip platform the sensor runs on and publishes its specification, and no outside model provider is named in the classification path: the company's own sensor and software produce the alert.

The privacy policy (Vayyar Imaging Ltd., updated 28 July 2025, covering the Vayyar Care sensor and the VC Install and VC Involve apps) names the other recipients by category: cloud computing providers for storage and backup, and customer relationship, chat and website analysis tools, none by name. It names Amazon, whose Alexa App, Alexa Device and Alexa Together Services receive fall alerts on the consumer route, and it permits the company to use aggregated anonymous information and disclose it to third parties without restriction. On the facility route, detections pass to the nurse call or care platform partner the operator buys through, such as TekTone or Austco, which then holds the resident's event record.

No subprocessor list, hosting provider, or position on whether data from installed sensors is used to develop the models is published. That matters for a deployment base that includes memory care residents who cannot consent for themselves.

Named deployments across several countries, with no measured detection performance behind them. Essex County Council in England reported eliminating hospital admissions and long lies after a large installation (January 2026); Sambrook House, a United Kingdom care home, published a resident safety case study (April 2026); an Ontario long term care provider piloted 122 units in a facility of more than 250 residents; and Coronado Heights Senior Living in Las Vegas is quoted in the company's June 2020 launch post, which also says the product runs in facilities across 15 countries and 4 continents. A distribution partnership with TLS dating to 2018 is described as having detected hundreds of real falls.

None of these carries a denominator. Falls detected say nothing about falls missed, and no sensitivity, specificity or false alarm rate is published for any installation. The council's outcome is reported in a case study rather than measured against a control period or a comparator technology, so the counterfactual is unknown. The launch post's "4X more accuracy than other automatic fall alert systems" names no comparison, method or source. The Vayyar Care page and the launch post cite no peer reviewed study of the product.

The evidence is named customers and case reports, and none states a method.

The design removes the most sensitive data outright, and the published terms say what is collected and kept but not what reaches a model. The sensor collects no optical data and cannot produce an image, which matters most in the bathroom, where falls concentrate and a camera is least acceptable to a resident.

The privacy policy (Vayyar Imaging Ltd., updated 28 July 2025) reaches the product: the sensor and the VC Install and VC Involve apps. It says the product collects only sensor based information that assists in detecting a fall, and lists what is transmitted: device identifier, time and date of a fall, Wi-Fi network name, geolocation and the room's name. Personal information is kept for as long as the person is a customer, or as law permits for legal claims. Aggregated anonymous information may be used and disclosed to third parties without restriction. The policy does not say whether sensor data is used to develop or train the detection models, sets no retention schedule beyond the customer relationship, and describes no de-identification standard. Safety engineering, meaning how a missed or false detection is handled as a safety event and what guards the model, is not published; the DCB0129 clinical safety work implies a hazard log that is not public.

The policy is written for the consumer route through Amazon's Alexa Together service; the company's business and customer terms, linked in the footer, may set different terms for facilities, where the partner's nurse call platform holds the event history.

A privacy document reaches the product, and no HIPAA or business associate status is stated. The privacy policy (Vayyar Imaging Ltd., updated 28 July 2025) covers the Vayyar Care sensor and the VC Install and VC Involve apps: what the sensor transmits, retention for the length of the customer relationship, disclosure categories, and deletion requests to privacy@vayyar.com. It is written for the consumer route through Amazon's Alexa Together service and does not mention HIPAA, protected health information, covered entities or business associates, and neither the Vayyar Care page nor the June 2020 launch post states a position.

The result is a service privacy document with no statement of status. The company's business terms and customer terms, linked in the footer, are the documents most likely to set a contracting position for facilities.

The channel makes the contracting question real. Facilities usually buy through nurse call manufacturers and care platform partners such as TekTone, Austco and K4Connect, so whether Vayyar signs directly with a covered entity, sits as a subcontractor to the platform vendor, or stays outside the health privacy chain is unstated. The DCB0129 clinical safety standard concerns clinical risk rather than data protection and does not answer it, and the absence of optical data lowers what is at stake without being a stated posture.

Controls are asserted with nothing independent behind them. The privacy policy (Vayyar Imaging Ltd., updated 28 July 2025) says the company uses appropriate technical, organizational and security measures and that it cannot guarantee information will not be compromised by unauthorized access to its servers. No security page, SOC 2 report, ISO/IEC 27001 certificate, trust center, penetration testing statement or vulnerability disclosure policy appears on the Vayyar Care page, the June 2020 launch post or the privacy policy.

The DCB0129 clinical safety standard achieved in January 2026 governs clinical risk management, not security: it says nothing about encryption, access control, secure development or breach response.

The device surface is substantial. These are connected sensors installed in bedrooms and bathrooms and integrated into nurse call systems, which are life safety infrastructure; device authentication, firmware signing and update practice, and how a sensor is provisioned and decommissioned are unaddressed. Installation and configuration are often performed by nurse call manufacturers and their distributors, so installer access sits between the vendor and the resident, and nothing describes how it is scoped or revoked. The data shows when a person is alone, asleep or in the bathroom.

No United States device position is stated, and the health regulatory regime the company engages with is British. Vayyar Care achieved DCB0129 in January 2026, as Care Home Magazine reported: the NHS clinical risk management standard for manufacturers of health IT, which requires a clinical safety officer, a hazard log and a clinical risk management file. DCB0129 governs how a manufacturer manages clinical risk. It does not say whether the product is a medical device or how it is classified, so it is not a regulatory position for the product. The device is also described as authorized by the Federal Communications Commission, which governs radio emissions, not medical claims.

The Vayyar Care page makes no FDA, cleared or medical device claim, and the product is sold for falls, presence, mobility, time at rest and bathroom visits with no vital signs claim. The page says the system analyzes trends to detect risk, and the June 2020 launch post described preemptive alerts to caregivers for signs of health deterioration. Neither is labeled cleared, and both rest on behavior a caregiver can check, such as a change in bathroom visits or sleep activity, rather than on a diagnosis the caregiver cannot see behind.

No device claim is made, and the product is scoped to falls, presence and routine changes a caregiver can check. The DCB0129 work is real clinical safety engagement that matters to NHS buyers, but it is not a device position in the United Kingdom or the United States.

Nothing is published on how the detection models are tested or governed. The Vayyar Care page, the June 2020 launch post and the privacy policy (updated 28 July 2025) carry no fairness testing, subgroup performance, calibration data, validation method or external audit.

The bias mechanism in radar is physical. The sensor reads energy reflected from a body, and the strength and shape of that return vary with body size, composition and posture. Wheelchairs, walking frames, hoists, adjustable beds and oxygen equipment put static and moving metal into the rooms where the product is installed, and bedding and clothing attenuate differently. Long term care and memory care residents are the people most likely to use mobility aids and to move unlike a typical development subject, and a model separating a fall from a controlled descent is exposed to gait variation from stroke, Parkinson's disease and frailty. How the system behaves with two people in a room, which matters for shared rooms and for couples at home, is not stated.

The DCB0129 clinical safety standard the company achieved in January 2026 would ordinarily require hazards of this kind to be identified in a hazard log. That is a clinical risk process rather than an evaluation of model behavior, and none of the analysis is public.

Every detection is routed to a person, and nothing stands behind the output. On the facility route the company says detection data is delivered into the central nurse call system, where staff receive, attend and log each alert, so a false alarm meets a person; a missed fall produces no alert, and nothing published catches it. On the consumer route the privacy policy (Vayyar Imaging Ltd., updated 28 July 2025) says the product sends fall alerts only to Amazon's Alexa App, Alexa Device and Alexa Together Services, "which are solely responsible for sending such alerts", which places responsibility for alert delivery with Amazon rather than with the company.

No detection warranty, service level, uptime commitment, indemnity or remediation position appears on the Vayyar Care page, the June 2020 launch post or the privacy policy, and no detection or miss rate is published, so an operator replacing pull cords or pendants cannot establish whether coverage improved. The company's business terms and customer terms, linked in the footer, are where a warranty or its disclaimer would sit. The DCB0129 clinical safety standard achieved in January 2026 requires a named clinical safety officer and a maintained hazard log, which is process accountability rather than recourse for the resident or the operator.

Named integrations with the nurse call and care platforms that receive alerts in senior living, and no record system. The company's own blog announces TekTone as a nurse call partner, and the named integrations include TekTone Tek-CARE and Austco Tacera nurse call, K4Connect K4Community, TLS Evolve, SenSights.AI, BLESEN in Australian aged care and eevi. The company says detections are delivered into the central nurse call system and states interoperability with any nurse call system and care platform. These are one directional feeds: detections written into a platform where staff receive and log them.

Choosing nurse call as the integration target suits the setting. In senior living the nurse call platform, not the medical record, is where alerts are received, escalated and audited, so detections enter an existing accountable workflow rather than a parallel console.

No electronic health record or care management system is named, no interface standard such as HL7 or FHIR is described, and nothing says whether a detected fall is written to a resident record as a discrete entry. Mobility, rest and bathroom frequency are trend signals, and a trend held only in an alerting platform does not reach care planning or incident review.

The sensor runs in the room it observes, and the vendor side is described only as outside cloud storage. The company says detection data is delivered into the facility's central nurse call system, which in senior living is usually installed on site, so part of the event path stays with the operator and its platform partner. The privacy policy (Vayyar Imaging Ltd., updated 28 July 2025) says the product transmits the device identifier, time and date of a fall, Wi-Fi network name, geolocation and the room's name, and that the company uses outside cloud computing providers for storage and backup, naming none and stating no region. On the consumer route, alerts go to Amazon's Alexa services.

No hosting provider, region, residency commitment, tenancy model or cross border position is published. That matters here: Vayyar Imaging is based in Israel, deployments are reported in the United Kingdom, Canada, Australia, Japan and the United States, and a United Kingdom public sector installation operates under rules on where personal data may be sent.

One hosted path, with its location neither committed nor stated, and the on premises sensor path described.

No price is published by the company, and how the product is sold can be established before contact. Facilities buy Vayyar Care embedded in nurse call and care platforms from partners such as TekTone, Austco and K4Connect, or through regional distributors, so the operator's price is set in that channel. Consumers reach it through Amazon: the company's consumer page and privacy policy (updated 28 July 2025) tie the home product to an Alexa device and an Alexa Together subscription, and the company does not say whether that service is still offered. Consumer prices reported on retail listings and review sites do not come from the company.

The coverage specification makes the unit count estimable. Range is published as 13 feet to the front and 6 feet 6 inches to either side, with a statement that large rooms may need more than one device, so the sensor count for a floor plan can be worked out without speaking to anyone.

Nothing states whether facilities buy the sensor outright, license it or receive it within a nurse call contract, what recurring fee attaches per sensor, or what the partner channel adds. No return figure is published with a method; fall reduction, avoided admissions and released staff time are argued without a cost figure.

Settings are named, and part of them has deployment evidence behind it. The company sells Vayyar Care for independent living, assisted living, long term care and memory care, and for private homes through Amazon's Alexa Together service; the June 2020 launch post says it runs in facilities across 15 countries and 4 continents. Deployment evidence exists for care homes and social care: Essex County Council in England, Sambrook House in the United Kingdom, an Ontario long term care pilot of 122 units, and Coronado Heights Senior Living in Las Vegas, with nurse call partners carrying the product in the United States, the United Kingdom, Canada and Australia. No study validates detection in any setting, and the home route has no named deployment behind it.

Clinical scope is deliberately narrow: falls, presence, mobility, time at rest and bathroom frequency, with no activities of daily living inference, condition prediction or vital signs. A buyer comparing it with wearable or multimodal monitoring products is comparing different products. How the sensor performs with more than one occupant, which matters for shared long term care rooms and couples at home, is not stated.