Write-up: seminar 6

Seminar 6 – Cycling and Transport Modelling
There were three sessions; two with longer talks on “Challenges” and “Developments”, and one with shorter talks on “Issues”. Some thoughts from the first two sessions are below; slides and audio from all sessions are here.
• Tim Gent made ‘The Case for Cycling in Large Scale Transport Models’ with an example from the Cambridge region, which unusually models cycling in its own right. This enables predictions of cycling uptake based on demographic shifts or changes in other modes, but isn’t good at estimating changes based on improved facilities or cultural/ attitudinal change. Many clients have lacked means, motive and/or opportunity to include cycling but social trends and new data might help change this.
- Should infrastructural preferences (when quantified) only be included in cost-benefit trade-offs, or should we also study and model a threshold effect for different groups?
• Helen Bowkett discussed why cycling is hard to model. The four-step model was developed to answer road-building questions and extended to cover public transport. For trip generation and distribution, we lack data on cycling and it’s harder to collect. For mode choice, choosing cycling confuses current models whose time/ cost calculations are set up for car and public transport. Route assignment algorithms select quickest or shortest routes, again inaccurate for cycling. Is it even possible to amend existing models to cover cycling?
- Is data really the big barrier? And is cycling inherently more difficult to understand and research than more dominant modes?
• John Parkin focused on more micro-level modelling for junction appraisal, and how these deals with cycling. Deterministic modelling uses equations based on Passenger Car Units (PCUs), which is problematic for two wheelers, particularly at higher flows (>20%). PCU values should be seen as variable depending on type of user and volumes. Micro-simulation offers more potential for studying bicycle movement, but more research is needed on cycling behaviour to feed into these models.
- In micro-sim modelling, how can we deal with the fact that in countries like the UK only a restricted demographic cycle, and their behaviour while cycling may not represent others?
• Herbert Tiemens spoke about the traffic circulation plan in The Hague, which involved the closure of parts of the city to motor traffic, and forcing through motor traffic to use the city ring. This was controversial, but re-routing motor traffic has had little impact on travel times and there has been some mode shift for journeys into the centre. The economic performance of The Hague has been good compared to other cities in the same time period, with air quality improvements.
- How can we better establish and use information about local economic impacts of encouraging cycling and/or restricting car use?
• Yaron Hollander and Aliasgar Inayathusein talked about work TfL is doing. They listed 10 challenges, including high forecasting error for minority modes, lack of historic data, and that cycling contradicts the travel time paradigm. One current project incorporates cycling into TfL’s Highways Assignment Modelling. The London Travel Demand Survey was used to distribute cycle trips, and network detail added (cycle-only links and minor roads). Trips were routed in the network based on signal delays, turn penalties, and attractiveness factors; routing was then calibrated against cycle counts.
- How can we help other areas to learn from such work in places like London and Cambridge?
• Paul Schepers spoke on modelling the unbundling- separating cycles from motor vehicles at route level – of Dutch cycle routes. He studied relationships between unbundling and bicycle use and safety, via a measure comprised of share of routes through residential areas, and presence of grade-separated intersections to cross distributors. More unbundling is associated with increased safety and slightly higher mode share.
- How can we better develop measures of cycling environments and infrastructure in the UK, for use in similar studies and in modelling?

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>