Hello Robot is sending its Stretch 4 mobile manipulator into the homes of older adults with mild cognitive impairment and early Alzheimer's dementia. The project started on 22 September 2026 and runs to May 2029 โ a clinical-study horizon, not a demo cycle.
Key takeaways
- The lead partner is the University of Illinois Urbana-Champaign, with the ClarkLindsey community as a second site
- The study is to involve 60 older adults together with care partners and professional care staff
- Initial testing runs at the university's McKechnie Family LIFE Home residential lab
- Eight-week evaluations in real care settings follow
- Funding comes from a National Institutes of Health grant of nearly $3 million
Three domains instead of a list of tricks
The study splits into three parts. Cognitive assistance covers reminders, guidance through an activity and locating objects. Social engagement means initiating interaction and maintaining contact. Safety covers mobility support and hydration reminders.
None of these functions is spectacular. All of them are measurable over time โ which is the point of the project. The robot is not meant to do one hard thing but to do simple things for eight weeks straight, in somebody else's home.
Hardware chosen for an apartment, not a warehouse
Stretch 4 is Hello Robot's current base model, priced at $29,950. The design is deliberately unusual: a compact wheeled base and a light telescoping arm instead of a heavy articulated one. That layout gets between furniture and does not need the manoeuvring space an apartment does not have.
It is the inverse of industrial humanoid logic. There, payload and cycle time count. Here it is footprint, quietness and whether the thing knocks anything over on the way.
The eight-week evaluation is the real test
Most robot presentations last anywhere from a few dozen seconds to two hours. An eight-week observation in a care setting measures entirely different things: how often the hardware failed, how often a human was needed, whether users simply stopped using it after a week. That kind of data rarely reaches a press release.
The grant details themselves โ the amount, the start date and the participant count โ come from The AI Insider. No confirming statement from the university or entry in the NIH project database is available so far.
Why it matters
Assistive robotics rarely gets funding long enough to measure abandonment by the user. A three-year project with 60 participants and eight-week observations is clinical-trial methodology rather than marketing. If papers come out of it, the industry gets its first hard numbers on how long an older person actually keeps using a robot at home.
What's next?
- The project is scheduled through May 2029, with results from the lab phase likely sooner
- The grant details come from a single trade outlet and need confirmation from the university or the NIH database
- At a $29,950 list price per unit, the study's scale stays limited to a handful of machines at a time
Sources
- Hello Robot โ Stretch 4
- The AI Insider โ Hello Robot Receives Nearly $3M NIH Grant for Assistive Robotics Research