DIY Segway – to motivate embedded programming?

Elektor magazine features in the current issue a DIY Segway called Elektor Wheelie. The system is build around a ATMEGA32 and has obviously a lot of mechanics. For sensing an ADXL320 (acceleration sensor) and an IDG-300 (gyro) are used. Looks like a fun project – and you have full access to the software (not sure what you really would program differently, perhaps one can tune it to get faster 😉

Perhaps it could be a platform to motive embedded programming – with clear real-time constraints, as it hurts if you fall off… Next term we are teaching digital system design and programming of microcontroller systems – should we get one for the lab? Someone willing to built it?

Information vs. Mobility, Percom PC meeting in New York

The PC meeting for Percom 2009 took place at IBM in Hawthorne, NY. Percom had about 200 submissions and many good ones – so we could compile an exciting program across the whole field of pervasive computing and communication. As one of three program vice chairs I have looked in detail in about 1/3 of the submissions that were application related. It is interesting to observe that research as a whole in the field becomes more major and at the same time more incremental. 

To me this puts up the big question in which domains will the new big innovations happen, what is the next trend after we have pervasive computing? There are luckily plenty of options, but at the moment it seems that there develops an interesting relationship between information, communication, mobility and energy. It seems that we can compensate mobility by information and communication and similarly we can reduce energy required by information available. One example is: if I know where things are (=information) I can reduce the effort required to find them (=mobility). Is there more to it?
Each time in the US – even in New York were public transport works quite well – one is surprise how alien it appears to many that it could be an option to take public transport on a business trip (e.g. there are no first class coaches on regional trains). Flying from Düsseldorf into Newark it was convenient to take the train to Penn Station in NY City and then an express train to White Plaines. If we would not have gone for a walk in the city we probably would have been equally fast as by car. With the again low gas prices in the US (less than 2U$ per gallon, down from 4 just a few month ago) I would expect public transport and small cars will not gain too much popularity – before the next rise in gas prices.

PS: it is amazing how many possiblities there are to serve coffee (and this is probably not one of the most environment friendly)

Hans Visited our Group, Issues on sustainable energy / travel

Hans Gellersen, who was my supervisor while I was in Lancaster, visited our lab in Essen. We discussed options for future collaborations, ranging from student exchange to joined proposals. Besides other topics we discussed sustainable energy as this is more and more becoming a theme of great importance and Pervasive Computing offers many building blocks towards potential solutions. Hans pointed me to an interesting project going on at IBM Hursley “The House That Twitters Its Energy Use“.

At the Ubicomp PC meeting we recently discussed the value of face-2-face meetings in the context of scientific work and it seems there are two future directions to reduce resource consumption: (1) moving from physical travel to purely virtual meetings or (2) making travel feasible based on renewable energies. Personally I think we will see a mix – but I am sure real physical meetings are essential for certain tasks in medium term. I am convinced that in the future we will still travel and this will become viable as travel based on renewable energies will become feasible. Inspiring example project are SolarImpulse (its goal is to create a solar powered airplane) and Helios (solar-powered atmospheric satellites). There are alternative future scenarios and an interesting discussion by John Urry (e.g. a recent article [1], a book – now on my personal reading list [2]). These analyses (from a sociology perspective) are informative to read and can help to create interesting technology interventions. However I reject the dark scenarios, as I am too much of an optimist trusting in peoples good will, common sense, technology research and engineering – especially if the funding is available ;-).

[1] John Urry. Climate change, travel and complex futures. The British Journal of Sociology, Volume 59, Issue 2, Page 261-279, Jun 2008

[2] John Urry. Mobilities. October 2007.

Poor man’s location awareness

Over the last day I have experienced that very basic location information in the display can already provide a benefit to the user. Being the first time in Sydney I realized that network information of my GSM-phone is very reliable to tell me when to get off the bus – obviously it is not fine grain location information but so far always walking distance. At some locations (such as Bondi beach) visual pattern matching works very well, too 😉 And when to get off the bus seems a concern to many people (just extrapolating from the small sample I had over the last days…).

In my pervasive computing class, which I currently teach, we covered recently different aspects of location based systems – by the way a good starting point on the topic is [1] and [2]. At We discussed issues related to visual pattern matching – and when looking at the skyline of Sydney one becomes very quickly aware of the potential of this approach (especially with all the tagged pictures on flickr) but at the same time the complexity of matching from arbitrary locations becomes apparent.

Location awareness offers many interesting questions and challenging problems – looks like there are ideas for project and thesis topics, e.g. how semantic location information (even of lower quality) can be beneficial to users or finger printing based on radio/TV broadcast information.

[1] J. Hightower and G. Borriello. Location systems for ubiquitous computing. IEEE Computer, 34(8):57–66, Aug. 2001.

[2] Jeffrey Hightower and Gaetano Borriello. Location Sensing Techniques. UW-CSE-01-07-01.

Just in time train schedule?

Thought experiment: if we have the same number of trains we have at the moment and we let them travel as we do at the moment – but without time tables (basically a train is always on time – it is there when it arrives – similar to today). Customers would have real time access to all trains and the system could provide estimates when a certain train is where – perhaps with a confidence interval and probabilities of connections and travel times (obviously with an understandable user interface).

Would this be a better or worse model of public transport?

… and by the way the coach I was in has the IP address and runs DOS 😉

Interesting MSc Dissertations at Trinity in Dublin

I have been at Trinity College in Dublin today as external member of the MSc exam board. While reading MSc dissertations I learned at lot of interesting things. Here are two pointers to technologies which I like to share (they may be useful in further projects):

SUMO – Simulation of Urban Mobility: An open source traffic simulation package

Small computing platform:

What do you decide in the car?

While waiting in Stuttgart in the lounge of the railway station I picked up a paper called “Auto-Bild” (the selection of magazines is really poor 😉 and I found an interesting news item in it.

KIA has done a survey (with over 2000 people) in the UK on decision making in the car. It appears that people use the time in the car to discuss major issues in their lives and that they make significant decisions during long journeys. I have not found the original survey from KIA but there are several pages that discuss the results, e.g. gizmag.

Some findings in short, people talked about/made descions: going on holiday (63%), buying a car (50%), moving (40%), getting a pet (26%), getting married (23%). The main reason for the car on a long journey being an effective environment for communication seems the fact the people are close together for a long time and no-one can walk away (41%). Also the fact that you have reason not to look the other person into the eyes, as you have to watch the street, was valued.

Thinking about it there it may also have to do with the function of space. A car puts people close together – in some case to intimate distances (up to 50cm) but defiantly to personal distances (50cm-125cm). There is a comprehensive overview by Nicolas Nova, Socio-cognitive functions of space in collaborative settings: a literature review about Space, Cognition and Collaboration (original reference to my knowledge is Hall, E.T. (1966). The Hidden Dimension: Man’s Use of Space in Public and Private. Garden City, N.Y.: Doubleday.).

This survey made me think more about the design space “car”. Recently two of my students – Anneke Winter and Wolfgang Spießl – finished there master projects at BMW looking into search technologies and user interfaces in the car. It seems there are a lot of ideas that can be pushed forward realizing Ubicomp in the car.