EUSAI (Ambient Intelligence) 2004
Eindhoven, The Netherlands
Ambient Intelligence is a step beyond Ubiquitous Computing. There is an explicit focus on individuals in an environment. As we look across Wearable Computers and this event there is a shift taking place. Tom Rodden, University of Nottingham, UK, put it bluntly - for 40 years our thoughts have been dominated by the PC, keyboard and mouse. This has actually been a diversion away from finding out what “will be our every day relationship with computing.” One of the most important changes is the social context and awareness that permeates the use of computers in every day life. The shift is profound: So far the individual has had to adapt to what the computer could do for the person; now with ambient intelligence the computer must adapt to the individual. These are implications that are only becoming evident.
The Technical University of Eindhoven, where the conference is being held, was showing an Electronic Paper Prototype with Visual Interaction with Windows. This is based on a large tablet, 36”+ diagonal that can be manipulated by a stylist or pen. Projected onto the surface are multiple objects which represent applications. These window frames can be manipulated with a small cursor box which is above the frame. The impressive aspect is that the objects, i.e., the windows of applications can be rotated just like a piece of paper. Even ordinary paper can be placed on the surface and a corresponding virtual paper can be rotated with these objects. The way that all of this works is that the objects are projected onto the tablet surface and the detection of what is happening on the surface is done by a camera which looks at the same surface. The objects on the surface are actually applications windows which are mapped as a texture on an OpenGL surface. Yes, this is not easy to explain without pictures. This is one of the most effective emulations of a desktop surface which uses computing windows we have seen.
This is the second European Symposium on Ambient Intelligence. The last one was sponsored by Philips and this one by the University of Technology, the Netherlands, also here in Eindhoven. There is no coincidence that much of this is centered here in Eindhoven, the headquarters of Philips. It has made a major commitment to Ambient Intelligence. There are 100 at the conference and most from Europe with a few from Asia. By every indication we have collected, the Europeans are well ahead of other parts of the world.
There are many definition of Ambient Intelligence and we offer some of the components:
This is quite broad but opens a new dimension on the relationship of the individual with their environment.
This conference is exploring in multiple tracks: Ubiquitous Computing, Context Awareness, Intelligence and Natural Interaction. We would broadly describe the presentations as
Tom Rodden, University of Nottingham, UK, is director of Equator. The project is described as:
What stood out in his presentation was the drive to take the technology into the “wild.” That is, he feels that to make the concept of ambient intelligence work it is essential to work with individuals. As a result, he described ways in which the technology has been used that were unforeseen.
His talk discussed five examples from Equator:
City was a museum project to show the city of Glasgow, Scotland. This mixes a physical visitor a museum with a VR Visitor and a web visitor. This brought a deeper visit experience and the video was interesting to see how the visitors interacted with others.
“Can you see me now” is an online game played in city streets. The objective was to locate the individual in the city. This uses GPS for position updates but there are errors due to masking. This significantly enhances the value of the game play and shows how sophisticated the players have become.
Ambient Wood is a children’s experience in the woods. It also uses technology in a number of ways to allow for exploration of the environment and biology. The woods were wired with WiFi. Children could probe the physical world and hand held devices were a part of this exploration.
An extension of the drift table, described at CHI, is the key table. This is a surface where individuals can drop items as they enter the house. The hypothesis is that the state of an individual’s emotion is reflected on how hard they deposit items on the table when entering the home. This is reflected in a picture on the wall which rotates based on the action one takes onto the key table. This picture frame had a dog in it. An unforeseen consequence of this test in one family’s home is that the family, and especially the mother, became very attached to the dog in the picture, and the emotions they felt were expressed in how the frame rotated.
After the presentation, the WAVE asked Tom why he did not show “more serious” applications of the technology. Yes, they do have such efforts and there are three under the Equator called digital care. One deals with Agoraphobia. Tom described that the involvement with health care as being an eye opener. One of the major issues is – what if the treatment is successful. This is not so easily dropped at the end of the experiment. As a result the art department has had to develop an ethics policy on how it will handle patients. In this case ambient intelligence is similar to drug care in a testing situation.
Ambient Intelligence is infrastructure associated with individuals. By implication it is every where, it is sensitive to presence and responsive. The WAVE saw some of the hard realities rise to the surface: where will the power come from for all these devices and what will it do for individuals?
No company is more committed to ambient intelligence than Philips. Some of its writings made available at the conference highlighted that focus and explained the position of the company.
The Institute of Pervasive Computing of ETH in Zurich Switzerland described how it would be possible to use RFID tags on a mass scale. Passive RFID tags can be deployed which are highly redundant and over large areas or object surfaces. They call their approach “super-distributed RFID tag infrastructure. SDRI.”
This allows for the discrete partitioning of the physical space. In the case of a regular grid pattern it is possible to have uniquely addressable cells and this can translate into a regular occupancy grid. This in turn, can support robot navigation. Thus, SDRI enables devices to interact with their environment. For example, it is possible to have an object move on a floor which has many such tags, and trace a virtual path on the floor.
“An SDRI can be regarded as a scalable shared medium with (almost) unlimited, independent, and highly distributed physical ‘access points.’ Further, if RFID tags support read-write operations, then they enable mobile devices to support a certain amount of data directly into the physical tags themselves.”
One mobile robot was developed using Lego Mindstorms. This was done to demonstrate the concept of an autonomous location-aware vacuum cleaner.
Overall, an innovative piece of work.
Marianne Petersen of the Center for Interactive Spaces, ISIS Katrinebjerg, reported on how we are seeing an explosion of domestic media from digital photos to home videos and more. She examined surfaces at home, such as tables and book shelves and categorized these. Home boards have information that is “one click” away. Then this was examined in terms of digital solutions. The analysis considered how both the physical spaces, i.e., the surfaces in the home, could be merged with digital media. The result was domestic hypermedia. This was termed – infrastructure for home materials. A MediaTable was outlined which became a collaborative and spatial organization of domestic material. This was actually displayed as a table. Also shown was a MediaWall. The talk was concluded with a next step that a MediaTable and connected surfaces would be placed in a real home for a 3 week trial.
A number of issues came up in conference, mostly during the questioning period. We summarize these:
This captures many of the thoughts we have had over the period of the conference. Some of the concepts, such as the MediaTable and the demonstrations in the Equator project, get close to making sense, from a consumer perspective. In fact, Tom Rodden, University of Nottingham said in his keynote that a number of companies have expressed an interest in commercializing the drift table. Yet, creating appealing products and profitable ones is not a given.
This is the only conference dedicated to Ambient Intelligence and thus a unique opportunity to evaluate the technology and status.
Ambient Intelligence (AmI) is a major step forward. Even if the net result of the focus of Ambient Intelligence is transitory before shifting to another technology, AmI causes institutions to change their thinking about the role of technology and individuals. This was very clear at EUSAI.
AmI is also about scale – the impacts of the technology are everywhere when it seeks to help the individual. The result is far reaching. The dimensions of AmI thus span Sociology, Psychology, Physiology and Well Being. But, it does not stop there. The technology also encompasses Philosophy, Evolution, the Life Cycle, Environment and Adaptation
Ambient Intelligence or Ambient Life?
We suggest that Ambient Intelligence is misnamed; it should be Ambient Life. The reasons are given above. The problem, as reflected at EUSAI, is that there is still too much emphasis on the computing or intelligence of the environments. Intelligence is one step beyond IT and IT is not where the thinking should be.
Thus, Ambient Life, as we are calling this, is about making technology with a higher purpose. It is a matter of thinking about products. It is at the center of where we go beyond the PC.