Report from the Secret Robot House

Last Wednesday, we organised an open robot house in Hatfield. This anonymous semi-detached house in a residential area of Hatfield was purchased by the University of Hertfordshire in May 2009 to do on-site research on robotic companions. If researchers were going to design for the domestic environment, might as well develop them in one!

They kindly let us see their work in progress, demoed 2 robots (Sunflower & Care-O-bot 3) discuss the research and have a bit of food. We invited Nicolas Nova, a Geneva-based researcher co-founder of the LIFT conference to come and give a talk about how science-fiction affects the design of robots and how roboticists are affected by those strong cultural messages.

Prof. Dr. Kerstin Dautenhahn who heads the research in the Robot House then presented the work done as well as some findings so far. Guests were then invited to ask questions and a debate developed around a few themes I thought I'd unpick a little.

1. Definitions & inspirations
Peter responded to Nicolas' presentation by pointing to religion and folklore as additional sources of myths that influence the design of robots. The group then discussed how a robot is even defined. There doesn't seem to be a standard definition on which research is done outside of a generalised idea of a smart product with some degree of autonomy & "intelligence". This of course applies to anything from a robotic arm or an Aibo dog. The lack of standard definition reminded me of the conflicts inside the design industry a propos the difference between design, art and craft, a well documented debate. It seems however, that by designing more and more robots, we help define what a robot is as well as identifying what makes us different from them. It seems to be an iterative process of discovering what makes us humans, what makes up "intelligence" and what we are capable of engineering.

2. Robots & the cloud
A lot of the processing power of Sunflower as well as its personality traits are now developed on board, making it a rather process intensive robot. James asked whether a cloud-based approach could be developed further down the line. This led to a conversation about migration of personalities in a home environment with several robots. How would you identify yours and how can you manage a distributed network of objects with their associated data & personalities. This relates very much to how we interact with everyday objects as well as our ability to manage data through our personal devices like tablets, mobiles, etc. There seems to be a real opportunity in developing platforms that make those transitions easier as opposed to developing very specific personalities for robots

3. Future scenarios & users
According to the research scenarios developed for Care-O-bot 3 the ideal user is an elderly person at home. A discussion developed around the usefulness for other markets and the limitations. As the research is ongoing in that area, there is much to do around how easily other types of users might accept these companions and how expectations can be managed. Will future elderly users who might be more at ease with computation in their homes react differently? Will they be more forgiving of the limitations of domestic robots? With the recent discovery of users rejection of Ri-Man in homes, the ethnographic waters are still rather murky.

All in all, there's still much to develop. As Kerstin underlined, research in robotics is a moving target, influenced and shaped by culture and research beyond the borders of the Robot House. As the project develops I'll try to keep you posted on the final versions of Sunflower and Care-O-bot and maybe even invite people for some more next year!

Attendees also blogged about the event here, here and here.

Posted by designswarm on Thursday, 16 June, 2011 /