The closing panel at Ubicomp dicussed the last 10 years of ubicomp and potential future direction (with regard to community and technology). On the panel were Gregory Abowd, Hide Tokuda, Lars Eric Holmquist, Eric Paulos, and Albrecht Schmidt. In the following I will just describe some of the points I raised in my short statement.
For me the first observation is that many ideas that were discussed at the first HUC99 (the first ubicomp conference started by Hans Gellsersen) - and were considered very speculative ideas have become common products and services by now (e.g. pocket bargain finder, predictive text input, mobile collaboration tool, mobile photo-sharing, location aware technologies). Here it is apparent that with regard to envisioning applications the conference has made impact. But there is a curious phenomenon: at the moment a device or service is available in the shop we do not recognize it as ubicomp anymore.
In some areas the complexity of the problems (when moving from the lab to the real world) has been underestimated - here context and context-awareness is a good example. If you realize the full vision it is basically solving AI. But nevertheless we progressed - there are applications for commercial mobile devices that do context and activity recognition - and it is just 10 years that we discussed this in a HUC99 in our paper on advanced interaction in context  - which was at that time really innovative! Context-awareness will happen - be patient :-) but its has to take into account: humans are adaptive, too.
The papers which had a large impact (based on citations e.g. check  on google scholar) seem more the papers that score high on novelty, even if the may lack scientific rigor.
For future directions I hinted some general directions (not necessarily my research directions):
- Implanted activity recognition and interaction (put the sensing and actuation into the body solves a lot of the problems … obviously it creates many new ones, too)
- Implantable persuasion and amplifying bodily experiences. Here I gave the example that we would be able to create a device to motivate you do sports by making your back hurt. I used this to emphasise that ethics will play a large role in the future…
- Prediction technologies (e.g. the weather forecast as an inspiration, forecasting traffic conditions, parking situation, restaurant business, costs, …) we will create systems that allo us to look up predictions (cost, quality of the experience, stress, time needed, etc.) for future activities (e.g. when choosing a restaurants, booking a travel, deciding on dating a person, making a business deal, accepting a position, ...)
- And finally I suggested that we will have fun with papers on privacy published now when reading them in 20 years :-) because our perception of this topic will change massively.
With regard to the community I made the statement that Ubicomp became the Starbucks of ubiquitous computing research - premium but based on the US idea of quality. Have you ever been in a Vienna coffee house, in an Italian espresso bar, or had tea in the middle east - it is very different. We lost some of the international spirit and we stopped arguing what good research in ubicomp is - this discurse should be started again!. Looking at the countries where ubicomp technologies come from (e.g. a lot from Aisia and Europe) we should again make a effort to more value the international diversity and the different styles and approaches in ubicomp research - scientific rigor is not the only axis to consider.
 Schmidt, A., Aidoo, K. A., Takaluoma, A., Tuomela, U., Laerhoven, K. V., and Velde, W. V. 1999. Advanced Interaction in Context. In Proceedings of the 1st international Symposium on Handheld and Ubiquitous Computing (Karlsruhe, Germany, September 27 - 29, 1999). H. Gellersen, Ed. Lecture Notes In Computer Science, vol. 1707. Springer-Verlag, London, 89-101. DOI= http://dx.doi.org/10.1007/3-540-48157-5_10