CityCyclingEdinburgh was launched on the 27th of October 2009 as "an experiment".
IT’S TRUE!
CCE is 15years old!
Well done to ALL posters
It soon became useful and entertaining. There are regular posters, people who add useful info occasionally and plenty more who drop by to watch. That's fine. If you want to add news/comments it's easy to register and become a member.
RULES No personal insults. No swearing.
not a support questionWith the world of whizzy computers, I was pondering how complex have the traffic models become and how much do the modellers understand them?
Are they an off the shelf product where you put in your layout, set some parameters and then leave it to runs multiple interations to work out the impact?
I find it all quite fascinating and would love to get my hand on a model to play with to see if minor tweaks make large differences in impacts. How do you become a modeller?
I have this question too.
There appear to be a few open source (free) transport simulation software packages available. Download and start playing.
as urchaidh said - there are quite a few open source packages, from my quick browse MATSim looks quite promising - http://www.matsim.org/open-scenario-data the example scenarios seem to work by using a street network (from OpenStreetMap) and population data, so where people live and work, shop etc then simulate those journeys.
Having a play with one of these has been on my things to look at list for a while, as I think it could be useful/interesting.
I think the hardest bit of this is working out the "plans".
There are really two types of transport model:
Microsimulation: These model the movement of individual vehicles as they travel around a road network. They use car following, lane changing and gap acceptance rules.
Macroscopic: These model traffic flow rather than individual vehicles, normally expressed as total flows per time period and averaged travel time per time period. In such models, all vehicles of a particular group are generally assumed to obey the same rules of behaviour.
There are industry standard modelling packages for each type. These include ARCADY, PICADY, LYNSIG (macro) and PARAMICS and VISSIM (micro).
These modelling packages have been under development for several decades by specialists at places like TRL. Modellers will be in control of the data input, but will have limited understanding of what is being done with that data to produce the results. These are fiendishly complicated mathematical models.
Does PICADY assume every individual suddenly gets into a car and drives directly into town when a new 1600 space car park is opened under a shopping centre?
jonty - Ha.
Actually, the traffic generating effects of new developments are normally assessed using TRICS. http://www.trics.org/
It lends a nice scientific edge to what is effectively reading tea leaves.
PICADY?!
Is that a spelling mistake or an omen?
I remember reading on twitter some remarks that no models capture bicycles properly, and that no models properly account for traffic evaporation, so the computer answer is always build for no less than the status quo of motor volume.
The Roseburn traffic modelling report explicitly said that it made no allowance for trips being diverted elsewhere or not made. That used the VISSIM model IIRC.
chdot - even worse, an acronym: Priority Intersection Capacity And DelaY.
There's a good article or two from a traffic modelling man turned part time critic somewhere out there on the internet. I don't remember who he is - but it ought to succumb to some careful searching. I think he wrote a book about this - and still provides modelling, but while also talking about the limits.
Will try to find...
My sense of modelling is that it never accounts for the stuff that makes for liveable streets. So my questions for modellers are things like...
Does the model account for the financial effects of the discouragement of people crossing streets by the most direct route (the answer will be no).
Does the model account for the discouragment/encouragement of cycling by the traffic density (I suspect 'no').
There's also one very very important feature I've seen misused in traffic models - which is when they are used to create an image of a traffic jam. In this scenario those queuing in cars take up many times the space in the image compared to those queuing on foot... so the image creates an utterly biased picture - a totally different picture from the same thing expressed in numbers (e.g. 10 people driving waiting 1 minute + 20 people on foot waiting 2 minutes - in pictures will emphasise the queue of space-hungry waiting vehicles).
Got it:
(scroll to the bottom)
I think Rachel Aldred also has some stuff on the inadequacies of traffic models when it comes to cycling.
Fundamentally modelling is a tool and while its operation is mathematically defined, its application is driven by ideology. If you close a street halfway along to cars and then model the traffic on it, you will discover entirely correctly it has become more difficult to get from one end to the other by car. The key is how much you care about how much more difficult it has become and how you trade off the other outcomes against that.
Rbrtwtmn - interesting article, which shares my scepticism. Thanks for sharing.
I can't remember where I read it, that conventionally one treats 'traffic flow' as being analogous to water in a pipe network, whereas we need to treat it as water that makes irrational decisions at every stage of its path.
Look at the temporary traffic lights just shy of the A720 at Lothianburn yesterday morning. Two lanes southbound on Biggar Road were available, yet the road capacity was halved because everyone was doing the British thing and queuing in a nice neat single line, and woe betide anyone who dared reach the front and attempt to 'cut in'. Traffic flow says plenty of capacity, but reality said drivers ended up blocking the Fairmilehead junction.
“conventionally one treats 'traffic flow' as being analogous to water in a pipe network”
Which also leads to ‘fewer lanes means traffic must go faster to get same amount through’
Or something.
“Traffic flow says plenty of capacity, but reality said drivers ended up blocking the Fairmilehead junction.”
It’ll be fine when it’s all driverless cars all obeying the same rules...
This is interesting:
---
Even when a proposed transport scheme is already at the heart of a heated debate, you rarely hear the objectors saying that they can’t follow the logic of the demand estimation work, or that the confidence in the outcomes seems farfetched. There is a surprisingly low number of Freedom Of Information requests about the process of developing models and forecasts.
It seems like transport modelling experts have somehow convinced the general public that there is no point in asking too many questions about their tools, maybe because they are too complex. This is unfortunate, since more challenge from those who are not close to the modelling work would impose better standards on this work. More scrutiny from the public itself would help us, modellers and planners, stop doing some things that we do just because nobody seems to notice.
---
From 'Who will save us from the misuse of transport models?', Hollander, November 2015, p.16
FOI request has been made... :-)
"It seems like transport modelling experts have somehow convinced the general public that there is no point in asking too many questions about their tools"
Perhaps, but now things are very much more 'we don't care what your projections are, they are based on you ALLOWING current traffic levels to persist AND encouraging an increase'.
These are POLITICAL decisions, model for reducing traffic instead.
It's the same old GIGO.
"It's the same old GIGO."
The aim of the FOI is to try and get evidence as to the nature of the GI which resulted in the GO. Rather than just inferring and/or labelling it as such without hard evidence, however politically justified that judgement might be, and however much 'we' might agree with it.
If government claims to be making "evidence based" policy and planning decisions, let's see the evidence...
You must log in to post.
Cycling in Edinburgh Flickr group
Video embedded using Easy Video Embed plugin
posts