Study of QR Codes in Southampton Bus Stops
by Christopher Gutteridge
Last year, an MSc student called Nick Gammer worked with data.southampton to study the value of QR Codes at bus-stops (linking to real-time data). There’s a few surprises in the results,
The key findings were:
The QR code trial was successful with increasing use over the life of the project.
Unprompted respondent comments were largely positive. The time trend of daily hit counts shows the project gained popularity over the time it was in operation, especially given book-marked hits were not recorded.
QR Code use was unaffected by the day of the week
There was little variation in average daily use and a One-Way ANOVA test confirmed no significant difference in average daily hits.
The QR code system was likely to have been beneficial to both regular, familiar, commuters and irregular, unfamiliar leisure travellers
Use was very consistent between weekdays/weekends and peak/inter-peak times. Also, survey results asking respondents whether they would be more likely to use the service on a familiar or unfamiliar route were reasonably evenly split (57% to 43% respectively). Furthermore, there was no significant difference in the change of acceptability of wait time between the two groups.
Scanning as opposed to entering the URL in a mobile browser was the preferred access method
Only 0.6% of hits recorded were through typing a URL
It is not beneficial to provide instructions on, or promotion of QR code use as there is no effect on uptake
Use of basic posters was higher in terms of absolute hits and footfall adjusted hits, however t-tests revealed not significantly greater for either data set.
It appears QR code posters placed at stops without a shelter receive higher QR code use
The average hit rate at stops without a shelter was substantially higher (28% greater usage) however due to insufficient without shelter stop numbers and footfall data this could not be formally tested.
There is substantial variation in use by area with the urban, university and interchange areas displaying much greater use that suburban areas.
This is true for absolute average hits per stop in given areas and even more pronounced when data is adjusted for footfall.
The presence of a display does not effect QR code use
Surprisingly the average footfall adjusted hit rate was not significantly higher for stops without a functioning display giving bus arrival times. This is supported by street survey data as respondents did not find wait time significantly more acceptable due to QR code use when a display was not available.
Use of the existing SMS arrival time service is low and could be redundant
None of the 67 street survey respondents used this service suggesting potential for replacement by essentially free and often real-time QR code provision.
The QR code system was easy to use
Eighty one percent of respondents found the system either very or quite easy to use.
Easy of use was not significantly influenced by smartphone ownership, however a significantly higher proportion of respondents aged over 40 found it more challenging.
Observed behavioural change was limited; the majority was in the form of utilisation of wait time.
Due to the methodology and high service frequencies the observed behavioural change was lower than previous studies. The only observed modifications were utilisation of wait time or going to a different stop. Further research is required.
System accuracy and reliability was adequate
Eighty eight percent of respondents believed the difference between their estimated and actual bus arrival times were less than the crucial 5 minutes. There were no system errors during demonstrations or from mobile survey comments and no vandalism occurred.
Arrival time provision through QR code use substantially improved the acceptability of wait time
Sixty five percent of respondents stated receiving wait times make their wait either much or a little more acceptable. There was no significant difference in the change of acceptability of wait time between respondents at stops with and without a display, frequent and infrequent travellers or passengers faced with a short and long wait.
QR code use lead to a valuable increase in feelings of safety
Thirty nine percent of respondents, mostly interviewed during daylight hours, felt safer after receiving wait times. A significantly higher proportion of female respondents exhibited positive changes in feelings of safety.
Potential patronage increases appear large although should be treated with caution
Fifty six percent of respondents stated they would be either a little or a lot more likely to use a bus as a result of QR code use. Previous studies reveal this is likely to substantially over-represent any resulting patronage change.
Knowledge of the difference between real-time and timetable arrival information is very poor and there could be substantial benefits from improving this knowledge
Only 10% of respondents knew the difference between timetable and real-time arrival times. Confidence in the system could potentially be greatly improved by informing passengers which times can be relied upon; confidence is an essential prerequisite for reaping the benefits RTI provision.
You can view the full report here: “An appraisal of QR code use to deliver bus arrival time information at bus stops in Southampton”
New and Automated Datasets
by Christopher Gutteridge
We’ve got two new datasets:
University Events Diary – this is aggregated once an hour from all over the university website and various other sources. It is going to be used to build a new university events website. Until then you can see the events on the pages for Faculties, Academic Units and Buildings. It (mostly) contains events which are open to the public.
SUSU Events – this is the same format, but contains events mostly restricted to our students. It’s pulled from their facebook page using the facebook API. Let me know if you want a copy of that script.
The Organisation Structure Dataset now pulls the structure and names from the university HR database every Sunday. The buildings occupied and homepages are still being hand maintained by me.
Data.Southampton Talk at SWIB2011
by Christopher Gutteridge
This is a video & slides from a talk I gave at SWIB2011.
-
View the Slides (PPTX)
Return on Investment
by Christopher Gutteridge
This is something I’ve been banging on about all year. We all tell people that open data is a Good Thing, but we don’t really provide any rewards for it. Search engines do — if you mark up your pages maybe they will do something with your vocab.org content, maybe not.
RSS feeds also provide an immediate and understandable ROI, maybe less than they once did, but people see their data getting used, and can play with it in tools from the moment it goes live.
I’ve been thinking that what makes a lot of sense is to start creating little widgets and websites that provide value on top of RDF open data. To this end I’ve created a little service which provides a building location lookup for your website, given an RDF document containing a list of your buildings. The code is available on request but is still a bit hacky. I’ve some ideas to make it really exendable and do some stuff to make it easier to customise. It’s written using ARC2, Graphite and some javascript.
I actually thing the demonstrations below are a really big deal. It’s one tool which can provide a useful service on top of the RDF from three different organisations. This is what I want the future to look like! These can be embedded in your own pages in an iframe, but it’s a bit icky loading 4 google map iframes in one page.
- Southampton [raw data]
- Lincoln [raw data]
- Oxford [raw data]
- Tsinghau (China!) [raw data]
- Your university? (let me know when you publish the data and I’ll add you!)
Also get in touch, maybe via the comments, if you would like to publish open RDF data about your organisation’s places. It’s dead easy and all you need is a UNIXish (Linux or OSX) computer, a free tool (Grinder), and my example configuration files for places in grinder. Given that you can edit your list of buildings (and sites, and rooms…) in a spreadsheet and turn it into RDF data in minutes. Cheap, easy, and with an imediate Return On Investment. This is how Linked Data should be.
I’m even thinking about making a Grinder web service, so you just publish a speadsheet, and a configuration file and we spit out RDF for you…
License Verification Link
by Christopher Gutteridge
A helpful chap named Glyn, from http://data.linkedgov.org/ pointed out that it’s hard to verify that the council really gave us permission to republish their bus data under the license we claim.
As a really simple and cheap solution to this, all pages on our site using this data, eg. http://data.southampton.ac.uk/bus-stop/HA030013.html now have a link at the bottom with the following text:
Bus stops, routes & live data provided under the Open Government License by the Southampton City Council ROMANSE office. To verify, see the Confirmation Page.
The important part is the ROMANSE project have provided a nice simple license verification page on their site at http://www.romanse.org.uk/sotonshared.htm which allows a 3rd party to verify in seconds that we’re telling the truth and are in agreement with the data provider. The above link is a page which just reads:
Southampton ROMANSE shared data
The ROMANSE Data on Bus Stops, Bus Routes and Live Bus Times is available via the University of Southampton Open Data Service http://data.southampton.ac.uk/bus-routes.html, Under the Open Government License http://www.nationalarchives.gov.uk/doc/open%2Dgovernment%2Dlicence/
I suggest that republishing someone else’s data, with permission, this is a good practice to establish and much cheaper than any other alternative.
Response to the Public Data Consultation
by Christopher Gutteridge
Here is a draft of what I’m planning to send to the Public Data Corporation Consultation. I’m posting it here first to see if anyone suggests any good improvements. I’ll send it in at the end of the week. The consultation closes on October 27th. If you care about public data in the UK maybe you should respond too!
* * * * *
Update: I went to hand this in today, and it seems that actually the consultation is a bunch of questions which rather threw me. I’ve done my best with it but it feels like answering “how would you like to pay to access the data?”. I am not very happy, it feels like there’s already a plan and they are just rubber stamping it. Sigh.
My basic take is that if data is required by government/council/public sector to run the country, then it’s required by the citizens to live in that country.
* * * * *
In brief: I believe the UK government should provide all public data with a license which allows free reuse (OGL), in formats which make it easy to work with, and identifiers which allow diverse data to be joined together in new ways. This will increase the wealth of all citizens and visitors to the UK. It will enable people in the UK to make better choices, and live better lives. The work begun by data.gov.uk makes me proud to be British and enables new kinds of benefits unprecedented in human history.
My name is Chrstopher Gutteridge. I am the Linked Data Architect for the University of Southampton. If it wasn’t for data.gov.uk this job title wouldn’t even exist!
I run the Open Data Service for the University of Southampton <http://data.southampton.ac.uk/>. This service was inspired by the UK government project, and has proved beneficial to our organisation with very positive support and feedback from as diverse sources as the Dept. V.C. for Education (Debra Humphis) and the head of our Catering services! By providing easy, open and joined-up access to information from the diverse parts of our organistation it means we improve the experience for our staff, students and visitors. The most beneficial tool using our data, to date, is a map of the university amenities <http://opendatamap.ecs.soton.ac.uk/> was not produced by our paid staff but by one our research students, who was keen to find a way to contribute.
I believe similar benfits and opportunities exist at the national and international level.
There’s two real benefits to the nation. The first is transparency. Allowing anybody to write tools using government data on things like crime, health, education, and other factors about quality of life or services is great — it helps people make informed decisions.
What is also a huge national asset is the fact that we’ve begun to publish catalogs of things in the UK, like postcodes, transport stops, postboxes, schools, parks, roads, etc. I see this as the digital equivalent of standardising UK plug sockets and domestic mains electricity.
If there’s a central way to identify, say, a road then any organisation from Google & Apple down to someone collecting a list of pot holes, they can all use the same code to identify the road. This allows organisations, or citizens, to later join up information from diverse sources to provide new value from existing databases. This is amazing and has so much potential. Not doing this is like allowing every train company to use a different train track gauge.
To use my own work as an example of how powerful this is. I collate information from University Catering on their coffee shops, from the timetable office on what teaching rooms we have, from our buildings and estates dept. on the ID Number, name, architect & year of construction, from the disibility office on the disabled access, and so forth. None of these departments need to talk to each other, but because all the other departments publish data using the building code number defined by buildings and estates I’m able to, very easily, join these up to create a far more useful resource: <http://data.southampton.ac.uk/building/85.html>
I’m very concerned that people might start having to pay to access this data. This will exclude the growing community who create computer and phone applications out of interest and enthusiasm and desire to make something to help people. It will exclude small companies, who can’t afford the risk.
I believe that unrestricted access, under the Open Government License, to all government and council data will make this a better country to live in for everyone.
Charging for government data would be like starting charging people to produce devices which use standard UK mains voltage — a regressive step which I believe woud do more harm than could ever be matched by the income generated.
I have used Linked Open Data to make the University of Southampton a better place. Please continue to do the same for the UK, Europe and the World.
Christopher Gutteridge – University of Southampton – http://users.ecs.soton.ac.uk/cjg/ – cjg@ecs.soton.ac.uk
* * * * *
2nd update; I just remembered this idea which I think is a sane compromise to kick start unfettered open data.
Hi, I already made a return to the PDC consultation but forgot to include something.
Christopher Gutteridge. University of Southampton Linked Open Data Architect but speaking as a private citizen.
Nobody yet has seriously investigating corporate sponsorship of open data. For example; getting every public loo in the country into a single database could be sponsored by Andrex. Anybody wanting to use this data would by able to freely use it under the OGL, but would have to credit “Andrex” in addition to the government.
This gives any individual or company unrestricted use of the data. Very large companies may choose to may a substantial fee to waive the requirement to credit the sponsor. I believe some companies call this a “white label” fee.
Admittedly some datasets would be controversial or hard to find sponsorship and it would be important to maintain public trust in the data, but there’s large amounts that this *might* work for and it ends with a better situation for the UK people.
The second any ‘hoops’ are required to to get to data (including API keys, license click-throughs etc.) then the power of it being open is massively reduced.
A very simple example of it being done right; The OS postcode interface to data allows you to view the data as an HTML web page:
http://data.ordnancesurvey.co.uk/doc/postcodeunit/SO171BJ
but because any tool can access and allow you to explore this data, you can view it in my own data viewer:
http://graphite.ecs.soton.ac.uk/browser/?uri=http%3A%2F%2Fdata.ordnancesurvey.co.uk%2Fid%2Fpostcodeunit%2FSO171BJ
or someone else’s
http://linksailor.com/nav?uri=http%3A%2F%2Fdata.ordnancesurvey.co.uk%2Fid%2Fpostcodeunit%2FSO171BJ&go.x=0&go.y=0
These examples are trivial but show software gaining direct access to the data. Imagine the frustration if most pages on the web made you sign an End User License Agreement before you could view them. It wouldn’t be as good, and nobody would really benefit.
I’m hoping that in the next year or two, things like arts festivals and museums may consider this sponsor-for-attribution model of financing so people could, for example, use the data from Edinburgh Fringe in any phone or web app, but the festival has already made its income by gaining sponsorship of the data, and the sponsors get their return as all users of the data have to attribute them in their apps. This allows an ecology of apps to be created rather than each data source only having a very few. This ecology will lead to innovation and improvements to the user experience. The more the application-creator has to pay for data, the less innovation we’ll see.
Updates and Automations
by Christopher Gutteridge
Org Updated
The “Org” dataset has been updated with data provided by our HR department on the new university structure, and the proper codes for everything. Where possible we’re using the two character alpha code (eg. FP for ECS) in the URI, but for the lower levels we’re using the full 10 charater code from the HR database, (eg. F7FP090000 for Web & Internet Science). This provides canonical IDs for all parts of the university, which is nice.
You can view the top levels on the organisation page. However, it now goes much deeper so I’ve added an option to let you view the entire org. tree.
Over the summer we’ve also added data on
- The location of student offices for each Faculty
- The location of the “deanery’ for each Faculty (although these usually are an entire floor, and our system doesn’t really do much with the concept of floors, yet)
- For faculties which have provided the data, a list of the buildings they ‘occupy’ (defined as controlling space in, rather than just having a member in that building). This has been provided by some faculties at the faculty level, and others at the Academic Unit level. We are happy to accept either. If we could complete this dataset it would be a very powerful resource.
- Where we can, we’ve added the homepage for that part of the University. We’ve not done this below the top two levels, but adding the data is very easy if people want it.
The spreadsheet in the get-the-data box on the organisation page is slightly unsual as I’ve set it to collect data we don’t show in the page. In addition to the URIs for the org. element and it’s parent element, and the label, it also includes a field for the code (FP or P2 or F5AF010000) — I figure that might be really useful for admin staff.
Update policy:
- Names of org. elements: This data belongs to HR and corrections should be sent to them. If they accpet it then I’ll update our data (please copy me the acceptance). Later, we’ll automatically source this data direct from their database.
- URLs for org. elements: data.southampton currently own this dataset, so it’s very easy to update. Just send us the correction to opendata@soton.ac.uk — I don’t anticipate much confusion so anyone is welcome to send us these.
- Space Occupied by org. element — This must be sent to us by an official admin for the group, unit or faculty concerned. If it’s a list for lots of org. elements, such as every group in a faculty, please get in tocuh as opendata@soton.ac.uk so we can send you a template spreadsheet to save us having to translate into the codes we use.
EPrints Automated
I should have done this ages ago, but the job got lost down the back of a sofa.
Every Sunday morning the EPrints datasets are now refreshed from the live archives.
Interlude: Seaside Data
by Christopher Gutteridge
So I’m off on my holidays, I’m going to be using the things I’ve been learning on data.southampton to use RDF data to power the events programme for a nice little arts festival in the town I grew up in.
I really like festival and conference data, there’s so many useful things you can do with it, and people just throw it into PDF or cruel abuses.
All the code will be reusable for future events…
Guest Post: QR Codes in Bus Stops
by Christopher Gutteridge
This is a guest post from Nick Gammer, a University of Southampton MSc student.
* * *
Just a quick note about a project I’m undertaking involving Southampton University’s Open Data Service. As the dissertation of my MSc in Transport Planning and Engineering and with the kind help of First Group and support of Southampton City Council I’ve placed QR code posters in over 40 bus stops around Southampton. These can be scanned using a smartphone which links to the next arrival times of all buses servicing the relevant stop, usually in real time as the majority of services now have GPS based AVL (Automatic Vehicle Location).
This information initially comes from ROMANSE, Southampton’s traffic control centre. However this is not mobile friendly and does not perform well on a smartphone. Southampton University’s open Data Service already skims this data, adding a map of the stop location, which is what the QR code links to. Without this service the project would be substantially devalued.
The project has been running for two weeks now with initially encouraging results; between 10 and 30 QR codes are scanned and link to this data daily. There is also a survey underway assessing users views of the service (amongst other things). I’ll put a quick note on the conclusion of the project at a later data.
Quick Update
by Christopher Gutteridge
IWMW: Dave & I have been away at IWMW 2011, learning and promoting the Southampton Open Data Service. A key message we’ve been pushing is to make sure the data providers get a return on their investment. If they don’t get a value to themselves or their users, why should they care? Very positive response from many people. I think the tipping point is upon us, finally!
Google Map: A few people have commented how the road on the Google Map goes right through building 85. I’ve put a report into teleatlas who provide that data and apparently a fix is in the works!
Bus Stops: I’ve added some neat new things to the sidebar [example]. A link to a fullscreen version of the bus times info, suitable for display screens (I think this would be nice in the NOC lobby), and also cut-and-paste code to add the bus times to your own pages, using an iframe.