This wiki is no longer maintained as of April 2012. Existing users may update articles, however user registration is disabled.
Community Assessment of Journeys to School
The experience children and their parents have of the journey school is important, it shapes the social development of children, it can influence their attitude towards (unstructured) physical activity and potentially influences the choices made by parents and children about how to travel to school. Local authorities in the UK and around the world have responsibilities for and influence over the built (and natural) environment that parents and children encounter in accessing schools. They must make decisions with scarce resources on how best to maintain and improve this environment. These decisions could be greatly improved with the consultation of affected communities to ensure resources are effectively targeted.
User generated data is a phenomenon that is giving rise to a number of innovations online and Ideas in Transit has identified a number of projects that are relevant to improving the experience of pupils and parents accessing schools. They have noticed a gap when it comes to the rating of individual sections of paths and roads as perceived by different members of the public (pedestrians, cyclists, people with push-chairs etc.). They have also noticed that there is no real way for these members of the public communicate with the correct authorities regarding these ratings and recommend changes based of local knowledge of the infrastructure.
We identify three main categories of users:
- Student and Parent Contributors. They are experts on their school run routes, which are typically well established and made at least twice a day. Parents and child will have different motivations for completing the map to a high quality. In a classroom scenario this will depend on the enthusiasm of the teacher. Some parents would be motivated by a sense of public spirit and the hope of change, others will not. The survey should be taken of whoever chooses the route: the student in the case where they cycle or the parent where they drive their child.
- Prolific Contributors. Those who will find and document routes they which they would normally not have cause to travel, contributing for the community's good and their status within that community. They will become experts on the system, but do not have specific expertise on the routes they contribute. An example of such a group is Living Streets, formerly Community Street Audit.
- School Management, Residents and Local Authorities. The output of this system is a view that allows the better understanding of the routes pupils take to school. This then becomes a tool for management and change.
It is proposed that data is captured by either by:
- An annotated printed map, in the style of Walking Papers, allowing low tech deployment - possibly with automatic techniques to import or assist with imports to the system's database.
- A website allowing people to specify their ratings for routes on a map interface, within a social community of users.
- A mobile application allowing assessments to be made en route and potentially geo-located, such as the FixMyStreet iPhone app.
There is a potential privacy/safety concern with tracing the route of children to and from schools. Processes will be needed to anonymise and control this information, with reassurance of this to the participants. Once the information is aggregated some of the privacy issues diminish, but remains sensitive.
A small initial trial is proposed using the Walking Papers format to gauge the kind of annotations people will make and their reaction to the task. A school needs to be identified for this and exercise completed quickly.
User's and their social relations are central to the system's architecture, they do not simply become a become a member of a predefined category. Instead they exist independently within a dynamic social network.
On first joining the community the user is presented with some archetypal images: a mother with a pushchair, a cyclist, a school going pedestrian. The selection of one of these creates a new profile in which lists a set of "sensitives" to various features, eg to a busy road or a steep climb. Each sensitivity is preset with a value, that the user can then tune to better reflect their preferences. As the community grows the archetypes and expected sensitivities can be derived from the existing users, using the cluster analysis mentioned later.
There is an open question about the dynamic nature of the features people will want to make assessments of and have sensitivities to. If this list grows beyond a small set of salient features, the simplicity of this system is at risk. Some features will relative to data that is already available via OpenStreetMap such that expected values can be set. For instance that steps make the route difficult for a pushchair.
A central function of the system is the ability to match similar users such that their ratings can be recommended to each other, as likely to be similar. This may first be estimated by the similarities of their sensitivities to features and transport type will likely be a major contributor to this, but not exclusively. Once suggested as friends, the users can make an explicit rate their affinity with the other user, based on whatever subjective judgment they would like to use.
We expect the resulting social network of stronger and weaker ties to yield useful and identifiable clusters. A user way wish to rate their affinity with the members of a whole cluster.
As a general design guideline, values may be estimated until they are explicitly set my the user.
Two functions of the system are then supported:
Rate a Route
Users may describe a route and highlight elements of it that are either positive or negative. The route may be traced by hand or uploaded via GPS. They will likely also highlight other features off the route, that influence their choice (and are likely to be negative).
The ratings made will concern specific features, for instance on the surface or safety of a junction.
Ratings made by other similar users may also be displayed to reflect the communities existing knowledge of that area. Allowing agreement or disagreement to be recorded.
Planning a Route
Using a routing algorithm and the weightings from the ratings of similar users, new routes can be suggested to to the user. The ratings used to calculate the route would be displayed on a map interface, allowing the user to re-rate them, or exercise some other social judgment about another user. This would allow the route to be refined.
The route planning facility proposed would have limited utility for parents and students, except when planning to change schools or at times when the regular transport infrastructure is unavailable. However, routing is an important component of a more general service that could be offered.
The diagram below represents the key screens of the web-based interface, as we currently envisage it. A new user is show first loosely identifying with a set of persona and setting up their profile and sensitives. The user may then describe a route or find a new route.
(23rd March 2010)
The Routino routing algorithm is being extended to enable subjective ratings to be made and accessed in generating routes. The Routino algorithm processes the OSM data to create an additional representation of the network that optimises the routing. We are working to extend the representation to allow features such as footpaths and turns to be identified and rated.
(4th May 2010)