In the field of network vulnerability and resilience, there is growing evident that network redundancy is a key determinant of network capability to withstand shocks. However, introducing redundancy to rapid public transport networks is very expensive and it is thus important to identify the network elements that will benefit most network performance in case of disruption.
It is crucial to invest in network redundancy without making redundant investments!
Together with my colleague from KTH, Erik Jenelius, we propose and demonstrate in a new paper a methodology for evaluating the effectiveness of a strategic increase in capacity on alternative public transport network links to mitigate the impact of unexpected network disruptions. For a given disruption scenario, a set of links is first identified as candidates for capacity enhancement based on (i) their initial saturation levels in terms of volume to capacity ratios, and (ii) the overloading in terms of increased saturation that occurs due to the disruption. Second, the effect of capacity increase is evaluated for each candidate link by comparing the disruption impacts with and without increased capacity. Based on the evaluation the most effective of the mitigation actions can be identified.
We applied the methodology to a case study of the high frequency public transport network of Stockholm. To evaluate the public transport system performance under varying conditions, BusMezzo, a dynamic public transport operations and assignment model was used.
The method presented in this paper could support policy makers and operators in prioritizing measures to increase network robustness by improving system capacity to absorb unplanned disruptions.
Click here for the full paper.
On Thursday, November 20, I had the privilage to present highlights from my research to MIT transit group led by Prof. Nigel Wilson and thereafter had the opportunity to disucss ongoing research activities with group members.
The seminar was entitled: “Unraveling and modeling the dynamics of public transport systems: Theory and applications”, where I briefly presented the transit operations and assignment model, BusMezzo, and its applications to service reliability and control, congestion and evaluation of increased capacity as well as service disruptions and the value of real-time information provision.
For a reduced version of the presentation, click here: MIT seminar 20112014 v1
The hEART2014 conference – the 3rd Symposium of the European Association for Research in Transportation was concluded today after 3 days of presentations and discussions in Leeds, UK. I have attended all three hEART conferences that have been organized so far and they have all proven a good opportunity to discuss ongoing works within an interesting research community.
I had two contributions on this conference. The first one is concerned with the evaluation of capacity increase in public transport projects. Many public transport investments are motivated by the need to relief congestion. However, conventional static assignment models that are used for cost-benefit analysis are not suitable for assessingi congestion related benefits as most of the negative conogestion impacts arise from system dynamics (e.g. reliability, crowding and denied boarding) and are therefore not reflected in average volume over capacity levels. Together with Jens West and Jonas Eliasson, we applied a new appraisal scheme for a metro line in Stockholm which takes into consideration the dynamic congestion effects and their variations. The presentation slides are available here: Appraisal of increased pt capacity hEART2014.
The second contribution is an attmept to infer the urban structure based on transport flow data. The availability of pervasive data collection facilitates the analysis of spatial and temporal distribution of activities based on people movements rather than based on land-use density proxies (as is conventionally done in urban planning and trip generation models). Together with Qian Wang and Yu Zhao, we formulated a spatial analysis technique to identify urban clusters and then classify them based on their temporal profiles with respect to incoming and outgoing flows. We applied this technique to Stockholm metropolitan area using public transport flow data. The analysis provide insights on the discrepency between the planning policy and the observed urban structure. This technique might be most valuable for anlayzing urban forms in mega-cities in emerging economies which undergo rapid changes. The presentation slides are avialable here: Urban clusters
It is common practice among public transport operators to regulate service departure times at several locations along each route. These locations are often called ‘time point stops’. Departure times from these stops are regulated in order to improve the overall service reliability. The operator has thus to decide which and how many stops along the line would be used as time point stops. This may seem like a simple problem but for a line with 33 stops there are more than 8.5 billion possible solutions (8,500,000,000)! Too many time point stops is difficult to operate and can slow down the service while too few will not be sufficient to prevent the deterioration of service reliability along the line. In order to address this problem of selecting time point stops with the objective to minimize total passenger travel times as well as operational uncertainty, an meta-heuristic optimization method was applied using a simulation model. The results show that the time point stops selected by the operator in the case study were worse than choosing them by chance! This method could be easily applied for other services to yield better reliability performance with very simple means.
Click here to read the full paper.