Cartogrammme pour les Votes du 5 juin

Appliqués pour la première fois au scrutin en Suisse, les cartogrammes tenant compte de la densité de la poputlation montrent la prépondérance des facteurs urbains et sociaux dans les résultats de vote. Source: Les villes ont massivement dit oui

Couverture du Flash Informatique

Petite fierté personelle de retrouver sa pomme en couverture du Flash Informatique de juin pour l’article de Patrick décrivant la plate-forme Moodle que nous mettons en place à l’EPFL.

Location Disculosure

Location disclosure to social relations:Â why, when, & what people want to share by the Place Lab people, explores whether and what users are willing to disclose about their location to social relations. The goal wat to understand the decision process and other factors that go into their decision that could inspire the design of better location-enhances applications and services. Their most notable result is that:

Participants want to disclose what they think would be useful to the requester or deny the request. We saw no evidence of participants intentionally blurring their location, i.e., disclosing something vague, to protect their privacy

Other very interesting findings are:

• Participants typically disclosed the most useful detail about their location (which is not necessarily the most detailed) or did not disclose their location at all. They often chose to disclose less specific information because they thought something less specific would be more useful to the requester and not because they were uncomfortable giving the requester more detailed information
• Participants chose to not disclose their location rather than merely blurring, which suggests that they were using the response to reinforce or communicate social boundaries.
• Who the requester was had the strongest influence on participants’ willigness to discolse.
• What participants were doing when they received a request appeared to have some effect on whether they would discolse something about their location

Self-Reported Positions are Credible

The Error of our Ways: The experience of Self- Reported Position in a Location-Based Game reports the experience and lesson learned from a location-based gamed in which players manually reveal their positions.

It appears that remote participants are largely untroubled by the relatively high positional error associated with self reports. The authors suggest that this may because mobile players declare themselves on their current trajectory (stating their intent) or behind themselves (confirming previously visited locations). Therefor, self-reported positioning my be useful fallback when automated systems are unavailable or too unreliable.

They state:

Analysing of the communication between and movements of street and online players revealed that the performance of GPS has a major impact on the game.

This goes a bit against the direction of CatchBob! in which we question the accuracy as being a factor that increase the performance. But maybe here “performance” is meant as a mix of accuracy, availability and latency.

The goals of the authors were to deepen their understanding of the human issues involved in using positioning systems. Secondly in the technique of self-reported positioning in its own right.

The first and rather obvious observation is that self-reported positioning provided excellent coverage and availability. Players quickly learning to use it.

Players appear to be anticipating time delay (human and technological latency). Declaring a few seconds ahead of themselves provides time for the system to respond with new information and maybe even for them to digest it before they reach the next decision point – a strategy that will avoid them waiting around.

There were also players declaring and looking behind their current position. Panning behind would often occur when a player did not manipulate the map for a while and so physically moved ahead of their last reported position. One reason for deaclaring behind was to retrigger clues or for the benefit of online players what had missed it.

Rather than reporting themselves to be a different place, the players are in fact reporting themselves to be at a different time

The authors suggests that explicitly self-reported positions (declarations) should be interpreted as deliberate acts of communication. The user intent is not captured by automated positioning systems.

Two potential limitations of self-reported positioning are that the mobile player has to know where they are and/or where they are heading, and that they may cheat. It also demands the constant engagement of the user in order to maintain an up to date position. It is therefor fair to ask to what extent technologies that are ubiquitous should also fade into the background and become invisible.

Digital Hotspotter

The Digital HotSpotter is my ultimate tool for my urban discovery. It has the same function as the special hazel sticks used by waterfinders. It allows to find sources… of wireless Internet access. Essential in worst case scenarios especially during trips and discovering urban areas when the coverage is unknown, protected or rare.

Ein Sonntag in der Innerschweiz

Financial Times' Cities of Dreams

Financial Time’s Tyler Brule came up with a list of The 10 top contenders for the title ‘City of dreams’.

Airports have been inspected, apartments assessed and neighbourhoods scrutinised – Fast Lane serves up its top 10 list of the world’s most liveable, loveable cities. The criteria demand that a city deliver quality of life across as many categories as possible.

The top 10 list is:
10. Montreal
9. Zurich
8. Palma de Mallorca
7. Munich
6. London
5. Stockholm
4. Sydney
3. Barcelona
2. Tokyo
1. Copenhagen

It is highly subjective of course but it matches very much the creative class’ criterias. I would add Geneva and San Francisco to the list, elect 2 cities and live in both of them by practicing frequent flying/train commuting.

Street Usage with Google Ridefinder

Via Mauro, a very nice map hack to visualize the paths of SuperShuttle buses in New York. Close to one of my possible future endeavor, collecting public transport usage data in Barcelona to create context aware applications.

I Started pulling down the XML for NYC every 5 minutes (the minimum update interval for ridefinder) and gathered about 1500 files over the past 5 days. Approximately 25-40 vehicles listed at any point, so at least 40,000 coordinates. The feed does not individually identify each vehicle, so it would be tricky to try and trace points together based on proximity and trajectory. And at five minute granularity, tracks would cross corners, connect over tunnels, etc.

The Impact on Collaboration of the Inherent Uncertainties of Positioning Technologies

Life on the Edge: Supporting Collaboration in Location-Based Experiences (CHI 2005, April 2005, Portland, USA), by Bendford (Mixed Reality Laboratory, University of Nottingham), Richard Hull (HP Lab in Bistol) and Jo Morrison (NESTA Futurelab) is an ethnological approach to an area I am very interesting in: the user’s experience of positioning technologies (or pervasive technologies in general) and more precisely the impact of their inherent uncertainties. I could find a lot of similarities between their experiment called Savannah and CatchBob! Themes I could bring back to CatchBob! are:

• How does the user interact with invisible sensing systems,
• How (un)awareness are the users of how they are being tracked (e.g. accuracy of the positioning),
• How do users address the difficulty of the system (e.g. when lost of connectivity or no position),
• How do users know and learn how to avoid and rectify the system’s mistakes (e.g. learning to detect the seams and cold spots),
• Do users detect uncertainty and what are the clues,
• How do users establish a shared context and coordinates their actions in moments of uncertainty.

In CatchBob! we thought of giving only rough positioning information or even inaccurate positions. My feeling is that in our game, users are more disturbed by the latency created by the communication system (some players even question if all the messages are being broadcasted) rather by the drift, jitter, lag and unavailability of the positioning system.

Some other pieces of information from this papers and about Savannah:

• Players halt when they encounter new information
• Players assume the the PDA is the sensing object, when in fact the sensor (GPS unit) is on their back
• System latency was also a factor in players’ difficulties; there could be a few seconds delay between a player’s PDA sending a position update, receiving new information from the game server and the player reacting and coming to a halt.
• There appears to be a considerable divergence between the player’s view off forming a group and the underlying system’s view, and this demonstrably can lead to serious confusion and frustration when players are unable to establish a shared context and act together.
• The system view of grouping is far more rigid. The system interprets multiple sources of data, which are often clouded by uncertainty and latency, and makes concrete and discrete decisions about when the groups have formed.
• When is it appropriate to work with absolute position and when relative?

Location-Based Applications at CRAFT

Nicolas and I wrote a 4-pages document that summarizes CRAFT’s research projects on Location-Based Services “lbs@craft“. It gives an overview of CatchBob!, Shoutspace and STAMPS. It also presents few results about CatchBob! as well as the issue we have to dealt with while designing those applications. It is not targeted to any special audience and will be used in the upcoming 2005 EPFL I&C Research Day.

In the last page, we talk about the various lessons we learning from the design of location-based services. I cover the current topics I am interested in pervasive technologies: Mixed approach to positioning, Indoor 3D positioning, Managing uncertainty, Latency, and engaging technology.