Kompaï in context
To put the Kompaï care robot into context, Robosoft is the main developer of service robots in Europe, probably the world. Unlike the android robots of science fiction and, lately, Japanese development, service robots are not constrained by having a human form and therefore come in many guises. They are focused on mechanical ways of achieving services provided by people that are time-consuming, boring and dangerous. For example:
- cleaning the glass pyramid outside the Louvre museum in Paris, which used to involve two nights to set up a crane plus two days of work for four mountaineers is now done in a day and a half by one person (on the ground) and a dedicated glass-crawling robot. And the result is better
- automated transportation in theme parks (cybernetic transport systems)
- multi-sensored reconnaissance and mine-defusing robots for the military
- robots that can refuel fleets of buses. This is an unpleasant and hazardous task but it’s a market that’s undermined by the human resistance factor – fleet managers who do not want to loose their fill-up-the-car perk!
For each problem Robosoft tackles they seek the optimal trade off between the minimum human intervention and the maximum autonomous behaviour of the machine. Each development involves solving new problems and matching them with the tried and tested modules that achieve many of the common movement and maintenance functions that a robot needs to do, such as to avoid obstacles or to dock itself for recharging.
Sharing the caring Kompaï
We noted the ‘birth’ of Kompaï in 2009 out of RobuLAB10, a ‘webcam on wheels’ that was designed to enable web-based, user-driven visits to sections of museums closed to the public. These two sitting were on a worktop as I went past.
I was interested to find out what has happened since, particularly after Robosoft took the bold move of making the Kompaï software open source a year ago.
Well, it seems that there are about 30 Kompaïs around the world being ‘worked on’ by various researchers. I picked up the clear message that Kompaï has moved from the prototype stage to being a ‘beta’. After further refinement it looks like there will be licensing opportunities for companies with manufacturing capabilities, if they are not lined up already.
Where would Kompaï sit in the care market?
In essence Kompaï is a replacement for the device end of a telecare system with added companionship features via the internet and the touch screen and webcam. Medication and appointment reminding is likely to be a standard feature of any setup and it looks as though it would be easy to give it the ability to suggest to users that they might like to play this game, or that, so it could have therapeutic benefits too.
So, going back to my opening question: Did what I saw convince me that the future of telecare is on wheels? Well, I’m still undecided, but I’m less sceptical than I was!
The answer will probably lie in economics rather than technology.
The essential factor will be whether mass production will bring down the cost sufficiently to change the return on investment calculations of such a robot for care provider organisations compared to the cost of sensor- and computer-based systems that would cover the same functions.
In the end it may come down to the relative costs and availability of the human resources needed to install and maintain the different systems and to train the users in their use.
However, Kompaï vendors should have one big advantage over sellers of boxes of sensors, controllers and standalone computers: its child-sized form, cartoon face and ‘top knot’ video cam undoubtedly give it the essential ‘cute factor’ appeal for the people who will have it in their house. They will therefore have a much easier time when it comes to generating media interest and marketing momentum.
Many thanks to the Kompaï team, François Hirigoyen, Patricia Commarieu and Yoann Pons for their warm welcome and time, and also to Robosoft’s CEO Vincent Dupourqué for making the visit possible.
Latest Robosoft video: Kompaï demo at Poitiers Hospital.