Anna Dumitriu and Alex May are currently resident in The University of Hertfordshire’s Robot House. The Computer Science department artists in residence are spending a week living in the house surrounded by research robots and are documenting their experiences here and making artwork in response to the situation. No-one has ever spent longer living in the house. The residency aims to raise public awareness around of The University of Hertfordshire’s work on the Accompany (Acceptable robotiCs COMPanions for AgeiNg Year) EU Project. There will be an artists talk (open to the public) about the residency and the My New Robot Companion project on 15th May at 1pm as part of the Computer Science Research Colloquium on the University of Hertfordshire Campus. There will be a Robot House Event with an exhibition, performance and a chance to meet the robots, including My New Robot Companion, which the artists working on throughout the week and the University of Hertfordshire’s famous robot KASPAR (pictured below). Sign up here for the Robot House Event, the house is in a secret location and you will need special directions.
We just gave a radio interview for Jack FM – Hertfordshire Radio about the open house event and our experiences living in the robot house. You can listen to it here: http://www.106jack.com/news/local-news/robot-house/ See their reporter Chris Hubbard’s face on our robot above.
UNIVERSITY OF HERTFORDSHIRE
COMPUTER SCIENCE RESEARCH COLLOQUIUM
"My 'New' Robot Companion"
Anna Dumitriu and Alex May
(Artists in Residence, School of Computer Science,
University of Hertfordshire)
15 May 2013 (Wednesday)
1 pm -2 pm
Hatfield, College Lane Campus
* * Lecture Theatre E350 * *
Everyone is Welcome to Attend
Refreshments will be available
The experiments continue. Today we participated in experiments around how close you would want a robot companion to come and how you would want it to approach you. I think Alex and I gave quite different results. I was happy for any of the robots, even Care-o-bot to come quite close, less than one metre away, whereas Alex felt he would prefer more space, especially if he didn’t know “what it wanted”. The Care-o-bot can move autonomously in the house, it glides across the smooth floor in quite a balletic way, using the full range of it’s directional wheels. This is quite strange as it appears to move backwards or sideways just as readily as it does forward, forward being where the nearest thing to a ‘face’ can be found. My only real concern with this was that sometimes it glided past me very close, so that I tucked my toes in. I was in stocking feet so a large robot driving over my toes might do some damage. The robot has a huge range of sensors but I asked the researchers if it would be able to detect my toes, where the sensors low down enough? The answer is no, the sensors can detect my ankles and are designed to allow the robot between small gaps, with “just a couple of inches” clearance, so it’s quite possible it could run over my toes.
As a treat we were given a demo of the Care-o-bot’s arm, a hugely elegant piece of robotic technology – like a much better version of the industrial lab robot arm in my studio (which is also lovely). I couldn’t help noticing a similarity between Care-o-bot and John Travolta is the movie “Saturday Night Fever” in the above picture though. I guess we all have a tendency to anthropomorphise. The gripper is impressive too. It took hold of a bottle of water and didn’t crush it. Apparently it’s sensitive enough to pick up eggs, the problem occurs as the programming isn’t complete yet so it doesn’t know how big things are when it puts them down so it can drop them.
Being artists in residence in the robot house in intense! We hardly have a minute to ourselves and instead are participants in a series of strange experiments. I must take the blame however, they asked us how much we were happy to offer ourselves as research subjects and I went on about the tradition in performance art of putting oneself into difficult situations, durational performance etc.
The day started with eye tracking experiments, when a robot interacts with me where do I look. The Care-o-bot followed me around the robot house and they tracked where I looked. Then it was the turn of Sunflower a cute square ‘droid’-like robot. The noted I didn’t look at the heads, which surprised the researchers. But robots don’t need heads, they are put there for our benefit, as decoration so to speak.
During the day Alex developed a vision system for our robot HARR1, so there we were tracking the vision of a robot – which is ‘attracted’ to things that move and ‘bored’ by things that don’t and the researchers were tracking our vision. HARR1 has a head, HARR1′s arms are the right length to be humanlike, this was not straightforward. The arms are just a collection of expensive servo motors, our collaborator Dr Michael L Walters spent quite some time creating lightweight spacers for the servo motors that extend the arms whilst still being possible for the motors to lift. Servo motors can break easily and the cheap ones used in maker projects would be useless for our purpose of creating a robot that can be used for extended period in art gallery settings without needing constant supervision.
We also had the opportunity to play with KASPAR’s new feature, to mediate conversations, to speak someone’s words. The idea being that it may be a useful way of interviewing children about certain issues (as yet undefined as far as I know) in preference to then talking to an adult. This is partly based on the work with children with autism who gain many benefits from supervised play sessions with KASPAR.
Well we are now ensconced in the University of Hertfordshire’s Robot House, in a secret location in a very ordinary residential area somewhere near the campus. The decor can be described as homely rather than high tech; the house is designed to study how people react to robots around them, and specifically the idea of robots as carers for older people, so the environment is clean and neat, but no single personality has come to bear on it, it’s not cosy, but rather sterile.
The ceiling has transparent panels of blue and silver dots about one metre apart, and lots of lights, these are to aid the robots in navigating around the space. The curtains are a mottled burgundy colour because patterns interfere with robot vision. The floor is absolutely smooth, so the robots can move around. I am writing this sitting (slightly uncomfortably) on a leatherette covered sensor pad so the house’s sensor system knows what I am doing. All the cupboards, draws, doors, even the bed and the toilet seat have sensors and our every move in the house is being collected for data by the researchers who work on the robot house. There is even a camera.
Today a researcher came to set up to KASPAR robots for us to use to communicate through between rooms. They can mirror our gestures and speak our words, not sure if they will use our voices or a voice more normal for the body of the robot, which looks somewhat like a toddler, if a quite scary one, at least until you get used to it. But one of the KASPARs was not behaving, so we will see it in action tomorrow now. I was reminded of something told to me by Artificial Life expert Inman Harvey once, when I showed him a robot I had built and apologised saying that it was not working at that moment, he said: “ah, broken and in need of repair, the proper state for a robot!”.
The most exciting robot we worked with today was our new Makerbot 3D printer, which we will use to print bespoke parts for our robot HARR1 (Humanoid Art Research Robot 1) who arrives tomorrow morning. We are looking forward to that! Today we printed a bracelet as an experiment.