What two years of ElliQ data actually measures

The ElliQ deployment in New York State ran for two years, reached approximately 900 older adults, and produced metrics the robotics industry has almost no vocabulary for. Ninety-three percent of participants reported reduced loneliness. Eighty-six percent said the robot improved their quality of life. Customer satisfaction: 4.9 out of 5.

Not one of those outcomes measures processing speed, actuator precision, or sensor resolution. Every meaningful result in this study measures something else entirely: how the robot made people feel. That is not a gap in the data. It is the finding.

The robotics industry runs entire conferences on locomotion, manipulation, and computer vision. It has almost no shared vocabulary for why one robot becomes a daily companion and another collects dust. ElliQ's results — deployed through a partnership between the New York State Office for the Aging, the Association on Aging in New York, and Intuition Robotics — are not explained by its spec sheet. They are explained by experience design decisions most robotics teams don't know how to name.

Those decisions are specific. ElliQ initiates conversation. It remembers context. It prompts physical activity and health check-ins without feeling clinical. Participants averaged 28 interactions per day, five days a week, roughly 28 minutes per active day. When 45 users showed up consistently for weekly live bingo — sending emoji reactions through their robots — that was interaction fit, not a technical achievement.

The largest share of usage — 38.4% — was companionship. Not utilities. Not health monitoring. Companionship. That function appears nowhere in a traditional requirements document. And the two-year longitudinal window confirmed the effect held: the 93% loneliness reduction did not fade after the first month. That kind of durability only happens when the experience is designed around human needs rather than technical capabilities.

"Engineering gets a robot into a home. Experience keeps it there."

Consumer robotics is entering a period where dozens of companion and assistive robots will reach the market. Most will be evaluated on what they can do. The ElliQ data argues for a different standard — signal clarity, interaction fit, failure transparency, and the experience dimensions that determine whether a robot becomes part of someone's life. Engineering gets a robot into a home. Experience keeps it there.