[ Home | Plans | Meetings | Members | Search | Papers | Links | CyberWear | LocoSoft]
The third International Symposium on Wearable Computing (webcast) was held in San Francisco on the 18th/19th October 1999, followed by the second International Workshop on Augmented Reality on the 20/21st October at the same location.
There were around 400 attendees at ISWC with a very wide range of subjects covered. This was a busy conference with a tight schedule. IWAR was much quieter in comparison. It had grown from a one day event to two days which it struggled to fill, though the attendance had grown from 63 to 175 participants.
The opening talk on Situated Computing gave an overview of current thinking (with a reference to Jim's '97 paper). It addressed some ubiquitous computing issues and applications (the intelligent medicine cabinet) and identified privacy as emerging as a BIG issue - along with personal tracking. Memorable quote - "we thrive on instability - wearable computing is truly disruptive!".
The first session presented four papers on Context Awareness. The first paper described 'Virtual Information Towers' - a neat way of accessing information as you move around, also mentioned Worldboard. This was a useful presentation which could help us organise our thoughts about data structures for location based data. The other papers described three alternative techniques for indoor navigation - RF tags; sensing devices(eg accelerometers - see also the TEA project); and hue maps - along with related applications.
The second session - Systems & Architectures - opened with Philip's Co-Modal Browser. They recognised that WAP is difficult to navigate and had produced an interesting configuration - palmtop + earpiece/mic. Also interesting was the MEX architecture produced by Nokia which was very similar in concept to our own EventManager. It also emerged during this session that there is a considerable user resistance to the Cyborg look.
Personal Applications was the third theme with an entertaining (but improbable) video showing a tourist application; an impressive augmented reality paper from Columbia showing further applications on their MARS testbed (rather cumbersome - but necessary for AR applications). This session closed with an interesting user studies paper on the electronic wallet - the three key factors being Security (offered by wearable); Vanity (nb Italians wear wrist and shoulder wallets); and Sentimentality (memory store).
The Gadget Show (Session 4) produced a stream of weird and wacky devices - from wax tablets to emotion sensing gloves. Of particular interest was a PS/2 chording keyboard, and variations on the pinger theme - it seems that IR pingers were initially favoured for their simplicity and low power consumption though problems with line of sight requirements have led to the increased use of RF.
Day 1 closed with Session 5 - Posters; Demos; & Reception - a very wide range of topics were covered here. notable was the new Itsy with an USB interface - and, no, they wouldn't let me have one and they're not going to market them - but they do intend to manufacture 500 so there's still hope. I found the nearest equivalent to our CyberJacket being made by a group of Lapps - a survival suit! Tried the MicroOptical glasses - very good for unobtrusively displaying a small amout of data. Our use of GPS and a compass with a wearable did not arouse much interest but the quad tree was considered novel.
Session 6 was about hardware - the inverted-F Antenna holds possibilities for mobile 'phone users. We also learnt that battery life is optimised by reducing peak loads (single threaded operation preferable) more than by reducing the active duty cycle and/or the idle load. Some useful context sensing stuff from Philips using accelerometers and resistive strips fixed to (bright red) jackets - future work here looks promising.
The next Session was probably the most useful :- Session 7 - Usability Panel. Here are a few extracts:
In Sessions 8 & 9 on Collaboration negotiating agents were presented; the effects of asymmetries in presentation of information; and there were two papers on video conferencing with some contextual sensing to make them more relevant! Once again a recurrent theme emerged known as (a variation of) the Martini effect - "right time, right place, right information".
The closing talk encouraged us to design wearables which would sense; think; and then communicate info which is timely, relevant, reliable, simple, situational, unambiguous. This should be provided in a fashion that is: unobtrusive, comfortable, convenient, and easy to use. We will need to combine:
Final thought from the platform - Wearable Computing will grow as telecomms industry grows - it has a great future
Press Story - Science News.
As the key issues in augmented reality are "registration, registration and registration", it was not really surprising that half the conference (seven papers) was devoted to tracking. Nevertheless it was an interesting 'Workshop' which will be upgraded to 'Symposium' status next year reflecting the growing interest in the field.
The First Session on Tracking comprised of three papers using vision; vision enhanced with magnetic prediction; and vision enhanced with inertia sensing to calculate the pose (position and orientation) of the users head. Prediction using inertia sensing appeared to hold the most promise.
The Second Session - on Mobility and Displays - presented papers on AR Conferencing (see also ISWC above); a wearable 'battlefield' system using dGPS; the UbiCOM wireless AR system using distributed StrongARMs (still in development - 15 staff, four year programme); and a well received paper on projecting AR images onto preformed 3D surfaces (e.g. bricks!). This session was particularly valuable as untypical - but potentially viable - applications were being presented.
For Session 3 we were back on Tracking with four papers addressing different aspects of calibration - all vision based. The methods included using an interactive system which referenced orthogonal objects; a 'card' system for AR conferencing (with a working demo); a predictive 'calibration propogation' method; and calibration using a zoom lens.
The final session - Applications and 3D Modelling - presented a virtual lighting system for rendering images with appropriate illumination; an industrial application for augmenting scenes with schematics with marker-less calibration; and lastly a system which merges X-Ray & Optical Images in real time.
The Dinner Talk concentrated on the privacy issues around wearable computers - it was based on a paper rejected for ISWC. From discussions with other delegates it appears that the freedom of information culture in the US has led to an unwarranted level of personal data being made available on the internet. The advent of wearables is seen to potentially provide "agencies" with an "amazing potential for invading privacy". For instance if we are connected to the internet and wearing a microphone and camera, our unwanted audio and video data could be 'sold' to a third party. The collation of this data from a number of wearables in the same area could enable the reconstruction of confidential exchanges.
In a 'breakout' meeting on human factors of HMDs the following effects were noted:- spatial disorientation; visual illusions; situational awareness (loss of); display latency; simulator sickness; sensor fusion. The problem areas were indentified as:- display fixation; info overload (less is more); standardisation of icons; comfort; accommodation of convergance. In general, HMDs are not received well - palmtop with camera attached proposed as better solution. Comfort and style need to be addressed by manufacturers.
The Posters & Demos included an impressive 'Border Guards' game produced by the Mixed Reality System Lab; as well as demos of the AR conferencing; a system using the microoptical glasses; and a camera driven from the movement sensors in a pair of Virtual I/O-Glasses (using flux-gate magnetometers).
The Closing Talk echoed ISWC with remarks such as "just say no" to HMDs - learn the lesson from VR work that acceptance is an important issue. Hand held displays need to be compared to head worn displays; and finally designs need to be user centred - so talk to the users first.
The material displayed is provided 'as is' and is subject to use restrictions.
For problems or questions regarding this web contact Cliff Randell.
Last updated: January 14, 2000.
ęCopyright Hewlett-Packard 1997-2000.