Mesoscopic model: cure for myopia?

Most people would not regard traffic engineering as sexy. Our memories of a visit to the zoo is full of cuddly animals and not the cafeteria or toilets. That does not mean the cafeteria and toilets are unimportant. We take the infrastructure in the city for granted. TCCS is a traditional organisation with traditional ways. The last legislative term was noticeable for the absence of cycle infrastructure. The Commonwealth Bridge crisis is changing TCCS but it is a painful and slow process. The outsourcing of the light rail planning may be symptomatic that they struggle with new ways.

A considered speech giving insights in transport engineering and TCCS by Ken Marshall at the Gungahlin Community Council meeting from 9 March 2022.

The future of traffic engineering

When you are considering building a road, you would expect the engineering calculations to include carbon emissions, as the emissions are an important long term cost to society. Traditionally, this is not the case. One would also expect that the alternate modes of transport would be considered. Infrastructure for light rail, cycling, walking has a cost but many benefits in cities where space is a premium and particularly walking and cycle infrastructure is much cheaper than infrastructure for cars, more space efficient with a great per square metre carrying capacity. Finally, the city is a business and walking and cycle infrastructure saves the city money as opposed to roads that need to be subsidised.

One would expect the engineering in TCCS to consider all these factors but they do not. The most recent trend with multimodal network planning is to consider all modes of transport – not just personal motor vehicles – and now the traffic modelling is finally following suit and considering walking and cycling as a legitimate form of transport.

While TCCS has many was to gather data and monitor car traffic (Bluetooth and sensors in the road), we have very few ways to monitor the traffic from pedestrians and cyclists. From an engineering perspective – evidence based planning – the engineers are flying with their eyes closed.

Climate change is in everyone’s mind. Carbon accounting is an attempt to look at the hidden cost of a carbon economy. A great idea but carbon accounting cannot be detached from traffic engineering and traffic modelling. When we are considering the “costs” of building transport infrastructure, we need to consider the lifetime costs. Investments in roads traditionally do not considered costs of carbon emission.

The Propensity to Cycle Tool (PCT) models a city and permits engineers and city planners to see which routes are most likely to increase the cycle participation and what carbon savings will come from building that infrastructure. The Propensity to Cycle Tool (PCT) model is an engineering approach to cycle network design:

  • Utilising real world data
  • permits the comparison of routes and
  • predicts the benefits to the city by scenario with before/after analysis.

It is strange that these models have been so long coming. The Propensity to Cycle Tool (PCT) has been developed in the UK since 2014 as a university research project funded by the UK Government. It is now being rolled out internationally, including New Zealand. However, the New Zealand example shows, that the Propensity to Cycle Tool (PCT) is not being adopted by government organisations, such as TCCS, as much as consulting companies (civil) and see the Propensity to Cycle Tool (PCT) as a service. The Propensity to Cycle Tool (PCT) is currently not available in Australia.

Where is traffic engineering today

In recent speech by Ken Marshall, TCCS, at the Gungahlin Community Council meeting on 9 March 2022, explained the current thinking to building roads for cars. The process he describes is a slow optimisation of road assets, weighing up the cost of investment with the safety benefits.

The quality of the recording was poor, which impairs the quality of the transcript. In a few sentences, the meaning of the sentence may even be obscured – in such cases the sentence will have been deleted. Ken Marshall spoke in a way which is typically for speech that feels rambling. However, the sequence of ideas are clear enough. We have added punctuation to make the text more readable.

Barton Highway roundabout

Take for example the Barton Highway roundabout. It is an extremely high-volume intersection. Consequently, there are many crashes that occur on that intersection but there has been previous significant investment and significant improvement. A lot of that “low hanging fruit”, in terms of preventing crashes, has already been achieved. There will be other sections where we can spend equivalent amounts of money and get greater effect than further treatments on an intersection like that (Barton Highway). Having said that, (the) intersection is key on the network and will continue to be in the long-term planning.

Ken Marshall, Gungahlin Community Council meeting, 9 March 2022.

What’s a model?

A model is an electronic version of the network that we can play with (and) that we can scenario test. We can change the input assumptions, and what a model is about, is predicting in the future (and) how the network will operate rather than understanding how it operates now.

Ken Marshall, Gungahlin Community Council meeting, 9 March 2022.

The network is shaped by historical data

A more traditional approach, (toward) which we lean, perhaps in the past have learned more heavily on, is the approach of taking what you might describe as lag indicators like crash stats, to evaluate how the network has performed in the past (…). And then use that analysis to identify weaknesses problems, deficiencies in the network, and devise programs of work that seek to address those targets, those identified deficiencies or problems: a bottom up approach or a lagging indicator forward (looking) approach to planning.

Ken Marshall, Gungahlin Community Council meeting, 9 March 2022.

The network is shaped by future expectation

The alternative approach, which we have always employed, but which TCCS is actively leaning towards in the current environment is a more strategic approacha top down approach.

That approach is useful in an environment where there is change or uncertainty, where the evidence from past performance might not necessarily be a great indicator of what the community needs from the network in the future, or indeed, where conditions are changing so that evidence from past performance is not a reliable enough indicator of what the future situation might require. That alternative involves

having a vision for what it is expected that communities will need from the network in the future, and then,

having a strategy that sets out a clear picture of what the transport network will need to look like,

what attributes it will need to meet those expected future needs.

We then develop programs of work that are driven by that strategy, that aim to shape the network into that future, rather than be made by the network performance in the past.

Ken Marshall, Gungahlin Community Council meeting, 9 March 2022.

Not one but both

Now in practice, of course, you did both those things. We have always done both those things at the same time. Ideally, (modelling) lived in the middle,

the past context: when you get initiatives for future improvement, or (mediation) of the network, that respond to problems that have been identified by that lagging indicator analysis,

the future context: respond positively to the strategic direction, then clearly, (to) an initiative that is a high priority.

So let me talk about (how) this process has traditionally informed programs. They are one of few inputs into the development of those types of programs. I suppose the real power (…) is that it tells us where to look. There are then a few other elements that need to be considered to determine where the expenditure will get the best benefit. That does not necessarily always mean the intersection with the highest number of crashes is the best place to spend money.

Ken Marshall, Gungahlin Community Council meeting, 9 March 2022.

Mesoscopic model

The ACT Government has developed and endorsed transport strategy, which is the guide and vision of what the network needs to become. The next step for us in implementing that strategy is the development of what we refer to as a Multimodal Network Plan. That is about drilling into the strategy, we want from the network and what bits of the network have got to do what to service all the modes – active travel, public transport, private vehicles – all of those modes working together, to (address) those aspirations what the community (is expected to) need from that network in the future.

There are a couple of feeds into that multimodal network plan. One of those (…) is the development of a mesoscopic model for downtown Gungahlin. That piece of work will happen this calendar year.

There are a couple of inputs that are needed in that model.

Let me maybe go back a step and explain these different types of studies: microscopic, mesoscopic and macroscopic.

Microscopic modelling is about how an intersection works or motor vehicles move through an intersection.

Macroscopic modelling considers employment and residences. What is the demand between the greater area or other town centres?

– In the middle, mesoscopic modelling is where we try and work out how the elements within that will work together – where the flows will be, and how the intersections will perform with those flows and how the modes of travel will work together.

We are working on (a tender) that goes out to market very shortly, (…) and it will be completed this year.

Ken Marshall, Gungahlin Community Council meeting, 9 March 2022.

Multimodal Network Plan

That will feed into the Multimodal Network Plan. And then that will help us to understand what needs to change about the way the network was originally designed, what is changed in terms of the land usages, and one of the keys to multimodal network planning is integration with land use planning, as well as integrating all the modes of transport together to get a holistic understanding of how we want that network to work. That analysis will give rise to a list of, potential mutations, improvements, alterations to the network that informs future capital works programs.

Ken Marshall, Gungahlin Community Council meeting, 9 March 2022.


Will the new mesoscopic model be a cure for a myopic traffic planning that has up until now disregarded active travel? There is no simple answer to this. Yes, a mesoscopic model will consider other modes of transport, however a model is as good as its assumptions and at this stage we do not know what those assumptions will be. The upcoming tender for the development of the mesoscopic model, that should be released in the next month, may provide further insight into this new model and the likely outcomes of its use.

Further, the mesoscopic model is not a clear statement from the ACT Government that they are choosing to pursue a network of grade separated bike paths. A piecemeal approach is quite consistent with the mesoscopic model. It is not known at this time whether the model will consider the psychological factors that make cycling appealing – particularly to women. Traditionally, psychology has not been all that important in traffic planning decisions – which is unfortunate as our behaviour and decisions are greatly affect by it.

Finally, we may have a long wait. It could be a number of years before we know whether this new approach will work for TCCS or whether we see a further car dominated infrastructure program justified by a myopic model weighted heavily by the status quo.


TCCS has not published a timeline for the development of the mesoscopic model and they are unlikely to do so. Best practice is to consider the development times for similar ventures to form a baseline (Thinking Fast and Slow, Daniel Kahneman). Developing a useful model takes time. TCCS does not currently have a mesoscopic model for cycling and are starting from scratch. As a baseline, we will use the development of the Propensity to Cycle Tool (PCT).

The development of the Propensity to Cycle Tool (PCT) started in 2015 and continued over a period of five years until 2019. We could similarly expect the TCCS mesoscopic model to have a 4-5 year development period, although initial results may be seen before that.

Timeline for the development of the mesoscopic model

  • 2022 first tender for the development of the mesoscopic model (announced March 2022)

  • 2023 first parts of the tool are working (prototype)

  • 2024 first projects as a result of the tools predictions

  • 2025 funding of the first projects

  • 2026 construction start of the first projects

    2026 mesoscopic model is now completed

  • 2027 completion of the first project (first stage)

“The work was initially funded by the English Department for Transport (DfT) to create the National Propensity to Cycle Tool for England (2015-2017, with further funding in 2018-19).”

About the Propensity to Cycle Tool, accessed 12 March 2022.

The development of the mesoscopic model is a welcome addition to the planning tools for the ACT but we will not see the benefits of the tool any time soon. It is also noticeable that most of the benefits will accrue in the next legislative period or even after that. It must be seen as a considered, long term solution and will do little to change much in the short term.

Bike paths not built Flemington Road and Gungahlin Town Centre, enlargement of Figure E2 Proposed Cycle Lanes and Share Paths, Ten Year Master Plan Trunk Cycling and Walking Path Infrastructure, Roads ACT, 2004.
2004 plan for bike path (BLUE) that were never built parallel to Flemington Road to connect Gungahlin Town Centre with Civic, enlargement of Figure E2 Proposed Cycle Lanes and Share Paths, Ten Year Master Plan Trunk Cycling and Walking Path Infrastructure, Roads ACT, 2004.

Leave a Comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s