Nobi
An AI fall detector shaped like an ordinary ceiling light, with imaging processed inside the light and not transmitted out
Nobi is an AgeTech company in Antwerp, Belgium, founded in 2018 by Roeland Pelgrims and others. Their product at first glance is just a cleanly designed smart ceiling light, but under the shade hide sensors and an AI computing unit that can detect a fall within seconds, automatically turn on the light, ask about the situation, and notify caregivers.
Features and application scenarios
I think this product design is one of the smartest in AgeTech in recent years: it hides the monitoring function inside an object elderly people would already accept. Getting a senior to wear a sensor wristband daily is hard, and installing an obvious camera in their room is harder — but no one resists having one more light on the ceiling. Technically, Nobi emphasizes that data is processed locally on the device and doesn't transmit footage out, simultaneously solving privacy concerns and network dependence. Besides fall detection, it also accumulates insights on sleep, activity, and daily routine. The official site claims care communities using Nobi have an average 55% reduction in falls. The company completed a €35M funding round around 2024, sells the product to 22 countries, and has bases in Belgium, the US, and the UK; Belgium's AZ Sint-Maarten hospital was the first hospital in the world to adopt it.
Best suited to residential care institutions, dementia-care units, and hospital wards; there's also a version for elderly people at home. Falls in Taiwan's long-term-care institutions are the No. 1 risk for accreditation and litigation, and this kind of solution that "doesn't need resident cooperation and starts working once installed" has a structural advantage in compliance rate over wearables. Price is the main barrier — European-designed-and-made smart-lighting hardware won't have a low unit price.
Main features
- Appears as an ordinary ceiling light, with built-in sensing and AI computing, non-intrusive
- Detects a fall within seconds and automatically lights up and asks the resident's condition by voice
- Real-time notification to caregivers or family
- Data processed locally on the device, footage not transmitted out, balancing privacy and network stability
- Sleep-quality, nighttime-activity, and daily-routine trend analysis
- Fall-prevention insights (identifying high-risk behavior patterns)
- Supports both institutional and home deployment scenarios
Common uses
- Real-time in-room fall detection at long-term-care and dementia-care units
- Patient fall-risk management in hospital wards
- Home-safety monitoring for elderly living alone
- Nighttime-activity and sleep-quality tracking as a basis for care adjustment
- Behavior-pattern analysis and prevention for high-fall-risk seniors
Key Features
- Appears as an ordinary ceiling light, with built-in sensing and AI computing, non-intrusive
- Detects a fall within seconds and automatically lights up and asks the resident's condition by voice
- Real-time notification to caregivers or family
- Data processed locally on the device, footage not transmitted out, balancing privacy and network stability
- Sleep-quality, nighttime-activity, and daily-routine trend analysis
- Fall-prevention insights (identifying high-risk behavior patterns)
- Supports both institutional and home deployment scenarios
Pros
- The product form solves elderly tech's biggest problem — acceptance
- No wearing, no remembering to charge — a key advantage for dementia seniors
- The local-processing privacy design is especially important for a highly sensitive space like the bedroom
- Has actual hospital-adoption cases and a concrete fall-reduction number
Cons
- High hardware unit price, and one is needed per room, so total cost rises linearly with scale
- Involves ceiling installation and electrical wiring, not plug-and-play
- Detection range is limited to the space the light covers; blind spots like bathrooms need separate planning
- The 55% reduction the vendor published is self-disclosed, without independent verification
Use Cases
- Real-time in-room fall detection at long-term-care and dementia-care units
- Patient fall-risk management in hospital wards
- Home-safety monitoring for elderly living alone
- Nighttime-activity and sleep-quality tracking as a basis for care adjustment
- Behavior-pattern analysis and prevention for high-fall-risk seniors
Editor's Note
Editor's note: This is one of my favorite product-design mindsets I've seen this year. The biggest failure reason in the elderly-tech field has never been that the tech isn't good enough but that 'seniors aren't willing to use it' — wristbands are complained about as ugly and clunky, and button-style call devices sit gathering dust in a drawer. Nobi directly sidesteps this problem: you don't need to accept anything new, you just changed a light. This understanding of user psychology is worth more than a few extra percentage points of detection accuracy. I think Taiwan's long-term-care-tech teams are well worth studying its product logic.
FAQ
Is it a camera? Will it invade privacy?
It has imaging-sensing ability, but the key is that data is processed inside the light and no video is transmitted out. This is a different architecture from an ordinary IP camera streaming footage to the cloud. Still, before adoption, the sensing principle and data scope should be clearly explained to residents and family — that's basic informed consent.
Compared to a wearable fall detector, which is better?
Each has trade-offs. Nobi's advantage is not needing resident cooperation and being suitable for dementia patients; the downside is it only covers the installed space, so once the senior leaves the room it's gone. Wearables' coverage follows the person but rely on the willingness to wear. In practice, many institutions pair the two.
Can it be bought in Taiwan?
Nobi currently sells across 22 countries, but the official info doesn't clearly list a Taiwan agent. Interested institutions need to contact the vendor directly to ask about Asia-Pacific sales and repair support, and clarifying this before procurement is advised.
