Using Commuting data from citizens to improve services

wotswattage

Has there been much data gathered on the populations (particularly Dublin) daily commuting habits?
Eg a mass study of a population considering where they need to travel from to get to work so a joined up transport network could be created.

Now maybe it does exist already I'm just not aware of it!?! What about other countries?

PhoenixParker

The census captures details of every morning commute which is available to researchers as the powscar data set. I worked with it for my PhD and it's a truly beautiful dataset.

Not sure how much it's used for service planning though. On the strength of the evidence I'd guess not at all.

wotswattage

PhoenixParker wrote: »

The census captures details of every morning commute which is available to researchers as the powscar data set. I worked with it for my PhD and it's a truly beautiful dataset.

Not sure how much it's used for service planning though. On the strength of the evidence I'd guess not at all.

Maybe I'm suffering from amnesia but I cant remember putting in details of whereabouts I worked in the census..?

EDIT: looking again at row actual forms you do input your workplace address!

PhoenixParker

Q19, 20 & 21 & 34

http://census.ie/the-census-and-you/the-census-form/

The form itself tells them where you're travelling from and other necessary personal details like age, sex etc.

skeogh82

The questions are flawed though - it only asks for the longest section of your commute.

For example I work on Mespil Road and I live in Northwood (Santry) and to get to work I drive to clontarf road dart station, get the dart to Grand Canal Dock, and then have to walk from there. All the census sees is i get the train - not that my commute is ridiculously un-connected.

Sure I could get the bus but that requires a 15-20 minute walk from Northwood to the bus stop, a 50 minute journey (even at 7am), and a 10 walk at the end - much slower than the my current but also inefficient route.

loyatemu

I'm sure google have pretty good data.

magicbastarder

as a tool for calculating movement patterns, the census is a fairly blunt instrument, i'd have thought. that said, it may be perfectly adequate, but more for work related commuting rather than general transport requirements.

PhoenixParker

skeogh82 wrote: »

The questions are flawed though - it only asks for the longest section of your commute.

For example I work on Mespil Road and I live in Northwood (Santry) and to get to work I drive to clontarf road dart station, get the dart to Grand Canal Dock, and then have to walk from there. All the census sees is i get the train - not that my commute is ridiculously un-connected.

Sure I could get the bus but that requires a 15-20 minute walk from Northwood to the bus stop, a 50 minute journey (even at 7am), and a 10 walk at the end - much slower than the my current but also inefficient route.

The census knows where you live and where you work and when you leave. The actual journey options you have can be calculated separately based on knowledge of the transport system and improvements can be modelled.

magicbastarder wrote: »

as a tool for calculating movement patterns, the census is a fairly blunt instrument, i'd have thought. that said, it may be perfectly adequate, but more for work related commuting rather than general transport requirements.

Work related commuting is the key transport requirement. It's the pinch point, solve it and everything else will typically be OK. Most leisure activities will centre around your home (known) and your work (known). They tend to be less static so what's accessible will be where you shop/socialise. Since they know where you live and work and where commercial/entertainment centres are other journeys can be estimated and modelled. For service planning an average across a population is what's needed.

Cookie_Monster

strava map cycling routes for those who use it.
not only useful but beautiful as well
they've started selling it to cities for planning purposes

http://labs.strava.com/heatmap/#11/-6.29345/53.33702/blue/bike

snickerpuss

The problem with the census question is you have to chose a specific time that you start your journey at. If you do shift work you might be in at 10am one week, 5pm the next. I don't understand why theres no "it changes/shift work" option rather than getting people to pick one answer which isn't really factually correct.

AlekSmart

snickerpuss wrote: »

The problem with the census question is you have to chose a specific time that you start your journey at. If you do shift work you might be in at 10am one week, 5pm the next. I don't understand why theres no "it changes/shift work" option rather than getting people to pick one answer which isn't really factually correct.

Leapcard,and particularly registered Leapcard, is now providing a fair degree of such commuting specific data,although once again,we remain uncertain as to what use it is being put.

Dublin Bus's Wi-Fi core data would also be of some use in establishing travel patterns,even down to deciding whether a specific journey was work related or not.

It is also worth noting that TfL in London is currently trialing a Wi-Fi monitoring system on the Tube,which in conjunction with Oyster data will fill many blanks in their understanding of the Commuting Human Mind....

L1011

It'd be a data protection nightmare but because the same firm provides close to all the bus and train WiFi and the system has GPS in it, very accurate cross provider tracking *could* be done.

Leap likely provides enough, though

Peregrinus

But Leap only provides data in relation to existing users. Google and the like have data that covers people who commute by car, by bike, on foot, as well as buses, trains and the Luas.

As for the data protection nightmare, I don't know that there's a huge problem selling aggregate data about journeys, routes, times, etc. For urban transport planning you need to know how many people need or want to travel a particular route at a particular time; you don't need to know who they are.

magicbastarder

Cookie_Monster wrote: »

strava map cycling routes for those who use it.
not only useful but beautiful as well
they've started selling it to cities for planning purposes

http://labs.strava.com/heatmap/#11/-6.29345/53.33702/blue/bike

the main problem with strava would be that it would favour leisure cyclists over commuter cyclists. can't see the actual figures, but on that map, the circuit of howth head seems to be as popular as the quays.

spuddy

The European Commission (& TomTom) have certainly looked at traffic levels using such data. In general terms, Dublin is the 4th most congested for its size, and at peak times it’s in the top ten worst cities of any size. They’ve also pinpointed why it compares so badly... (surprise, surprise)

“The shortage of mass transit facilities around Dublin has led to increasing road congestion and high associated economic and environmental costs. Dublin was the fourth most congested city of fewer than 800,000 people in the latest TomTom traffic index, with an overall congestion level of 38 % ( 57). If only peak morning and evening hours are considered, the congestions index surges to 81 %, ranking Dublin as the ninth most congested city of any size among more than 200 cities monitored by the index (graph 3.3.3).”

http://ec.europa.eu/europe2020/pdf/csr2016/cr2016_ireland_en.pdf

Banjoxed

spuddy wrote: »

The European Commission (& TomTom) have certainly looked at traffic levels using such data. In general terms, Dublin is the 4th most congested for its size, and at peak times it’s in the top ten worst cities of any size. They’ve also pinpointed why it compares so badly... (surprise, surprise)


http://ec.europa.eu/europe2020/pdf/csr2016/cr2016_ireland_en.pdf

Official Ireland's preference for commissioning endless reports to ensure politicians don't make the 'wrong' choice is the root of that statistic.

magicbastarder

are there any significant holes in the data available? i.e. what might be missing which prevents better public transport planning (as opposed to better public transport investment!)

suicide_circus

snickerpuss wrote: »

The problem with the census question is you have to chose a specific time that you start your journey at. If you do shift work you might be in at 10am one week, 5pm the next. I don't understand why theres no "it changes/shift work" option rather than getting people to pick one answer which isn't really factually correct.

Yep, there is no scope for capturing the hundreds of thousands of people working shift/flexi-time etc.

bigredone

Google have all this data unless you turn of location tracking on your phone.