When I picture the robot carer that might look after me when I’m old, I see something vaguely human – an endlessly patient android that can make dinner, help me out of bed, and pretend to laugh at stories it’s heard 300 times while it secretly scrolls through whatever the robot equivalent of Tinder is on its internal smartphone. The reality already rolling into people’s homes is less dramatic – though probably far more useful.
Take Barcelona resident Irene Veglison. She hasn’t danced in more than two decades. Three months after Sandi moved in though, she was swaying along to music again. Sandi is not a live-in carer or unusually energetic new flatmate, as you’ve probably guessed. Instead, it’s a 1.35m-tall robot with a screen for a face.
Sandi reminds Veglison to take her medication, tells her about appointments, wakes her in the morning, and can connect her to a social worker in an emergency. Sandi is, in fact, one of 600 robots deployed in homes and care centres through an EU-backed Barcelona programme for people experiencing the early stages of cognitive decline.
Okay, so it’s not quite The Jetsons. But it offers a much more believable glimpse of robot care – handling small, repeatable jobs, spotting when something might be wrong, and keeping real people within reach.
The need is certainly there, and while robots can’t fix that shortage, they could help people stay independent for longer – and give overstretched carers more time for the work that only a human can do.
Companionship comes first
The first robot to care for you might not have arms or legs. It could look more like ElliQ – a tabletop companion with a swivelling head and touchscreen, designed specifically for older adults.
ElliQ can start conversations, remember personal details, suggest gentle exercises, offer medication reminders, and help its owner exchange photos and messages with family. It can even play bingo.
Unlike Alexa or Siri, which normally wait to be summoned, ElliQ is designed to start conversations based on someone’s routine and previous interactions. That could make it feel less like a talking appliance and more like company – even if nobody is going to mistake its lamp-like head for a human face.
A robot cannot replace a friend or relative. It can, however, fill some of the long, quiet gaps between human visits. It might encourage someone to move, eat, drink, take their tablets, or call a loved one. Those sound like small jobs – right up until they help somebody carry on living safely in their own home.
The evidence is promising, but still limited. A 2025 review of 40 studies examined older people’s attitudes to care robots, cognitive support, and help with everyday activities. It concluded that more capable machines – plus help with buying and maintaining them – would be needed before they could become widely useful.
Keeping watch
A care robot could also put a friendly face on the sensors dotted around a future smart home. Cameras, motion detectors, and connected health devices might help it learn what a normal day looks like – then notice when something changes.
Maybe its owner hasn’t entered the kitchen by lunchtime. Maybe they’ve repeatedly missed their medication or don’t answer after a suspected fall. The robot could check in, contact a relative, or bring a professional into the room over video.
Some of that is already possible, but the details matter. Barcelona’s robots can connect users to social workers, who can access a camera during an emergency. The city wants future versions to detect danger automatically – but that is the plan, rather than a feature working today.
ElliQ is similarly clear that it is not a medical device, which is obviously a very important distinction. A missed meal could signal illness, or it could mean someone popped out for lunch. A robot can spot the change – but a person still needs to work out what it means.
An extra pair of hands
Talking is the relatively easy bit. Homes are messy places filled with stairs, rugs, pets, clutter, breakable objects, and people who move without warning. A robot has to navigate all of that, pick things up safely, and recover when something goes wrong.
Labrador Systems’ Retriever takes a simpler approach. It is essentially a very clever self-driving trolley, built to carry household loads and collect prepared trays from fixed locations. Its adjustable deck could bring food, drinks, or medication to someone with limited mobility.
That is much less impressive than a robot nurse – until carrying a cup of tea across the room becomes difficult. Even so, Labrador’s own website describes its specifications and product plans as targets. The robot is limited to one floor, and the company recommends adult supervision.
More human-shaped machines promise greater flexibility. The bipedal 1X NEO is taking deposits through an early-access programme. It is designed to open doors, fetch objects, tidy rooms, and tackle laundry. NEO looks much closer to the robot helper many people have imagined. But 1X says early units will have only basic autonomy. If it encounters an unfamiliar job, a remote 1X expert can supervise it at a scheduled time.
That’s impressive – and a reminder of how far there is to go. Folding a towel with remote help is very different from lifting someone out of bed. Bathing, dressing, and toileting are exactly the jobs that could relieve the most pressure on carers, but they are also the ones where a mistake could be painful, dangerous, or deeply undignified.
Why people still matter
A robot that watches, listens, and remembers what happens inside your home creates some obvious privacy problems. Its sensors could spot a fall or find your glasses – but also capture private conversations, medical information, and visiting family members. With remote assistance, another person might temporarily see through its cameras.
There are emotional questions, too. Someone could form a real attachment to a companion that belongs to a company, depends on remote servers, and might change after an update. A person experiencing cognitive decline might not understand what is automated, what is recorded, or whether that friendly voice understands anything at all.
None of this makes robot care a bad idea. It means consent, simple controls, strong security, clear data rules, and human backup matter as much as clever AI and dexterous hands. Cost matters too – care robots will not help many people if they remain expensive toys for wealthy early adopters.
The robot that eventually cares for you probably won’t, then, replace one person, job for job. It might not even be one robot. A tabletop companion could handle reminders and conversation, sensors around the home could watch for trouble, and a mobile machine could carry things or tackle simple chores. Human carers, relatives, clinicians, and emergency services would step in when judgement, touch, or responsibility mattered.
That’s all far less exciting than an autonomous android nurse, but considerably more useful. The best robot carer might be the one that handles the dull, repetitive work, spots trouble early – and knows exactly when to call a human.
