Just like the wheelchair which has been studied to death by academia but never developed as a practical solution, eldercare robots have also been studied heavily and resulted in nothing marketable. Kompai is the perfect example of this.

Read the abstract:

http://www.aaai.org/Papers/Symposia/Spring/20...07-012.pdf

"Abstract
Teleoperated from a distant location, a mobile robot with some autonomous capabilities can become a beneficial tool in telehealth applications. Assistive technologies for telementoring in homes constitute a very promising avenue to decrease load on the health care system, reduce hospitalization period and improve quality of life. However, design issues related to such systems are broad and mostly unexplored, but with very few systems currently available commercially. Mobile robots operating in home environments must deal with constrained space and a great variety of obstacles and situations to handle. This paper presents the interdisciplinary design methodology followed to develop Telerobot, a telepresence assistive mobile robot for home care assistance of elderly people. Using field trials with existing platforms, focus groups and interviews, initial requirements for the new mobile robot platform with its augmented video user interfaces are outlined."

Spencer did his focus group about 10 years before and when he went into the home trials they interacted with the elderly and he knew in advance that the CareBot would fit through the doors.

"Preliminary Studies
Three types of preliminary studies were conducted in our project: telerobotic systems in home environments; focus groups with healthcare professionals and elderly people; interviews with system users to model the health information architecture. The objective of these studies was to gather more information to come up with good initial specifications for the telerobotic system."

"Therefore, we conducted a pilot study to evaluate two conceptually different user interfaces for teleoperated mobile robotic systems, with trained and untrained operators (Labonte et al. 2006). This study aimed at identifying locomotion and structural requirements for the new robotic platform, as for user interface requirements for improved efficiency and security of novice operators of mobile robots. The objective was to identify critical elements that must be considered in the design of such system, before considering conducting an in-depth study (e.g., statistical analysis with a probabilistic sample)."

It's a paper about a preliminary study, not an in home alpha trial of a functioning system.   It seems to use predecessors of AVA, which is still a developmental platform.    IRobot's current medical telepresence  platform is for hospitals and it's so expensive they can't sell it, they can only lease it to selected customers, apparently for extended alpha testing.  It is not a marketable product.

GeckoSystems has the only eldercare robot that has been thoroughly  Alpha tested and is ready for final development and Beta testing. It could be on the Japanese market by 2015, the first mobile service robot to qualify for 90% coverage under Japanese nursing care insurance.

GeckoSystems also has the only wheelchair with a marketable AI robotic navigation system.  With adequate funding or the right partner or licensee it will be on the market in a matter of months.