Gave on January 18 a seminar at C2Smart together with Panchamy Krishnakumari, a colleague in our Smart Public Transport Lab. We were hosted by Prof. Joseph Chow.
You can watch the seminar by following this link.
Our seminar was entitled Capacity Allocation for On-demand Services, Demand-anticipatory Operations and Analyzing Demand Patterns. On-demand transit has become a common mode of transport with ride-sourcing companies like Uber, Lyft, Didi transforming the way we move. With the increase in popularity for such services, the fleet needs to adapt according to the demand and passenger demand needs to be predicted. In the seminar, we presented our work on capacity allocation for on-demand services, demand-anticipatory operations and analyzing demand patterns using spatial-temporal clustering.
Panchamy and I gave the seminar following the TRB conference in Washington DC and a project meeting with WMATA (Washington Metropolitan Area Transit Authority).
Crowding in public transport can be of major influence on passengers’ travel experience and therefore affect route and mode choice. The impact of crowding on passenger choices has been estimated in many studies by means of stated-preferences choice experiments. Respondents are then asked to make hypothetical choices under a range of scenarios based on which choice models can be estimated, including quantifying the impact of on-board crowding on route choice. This results in in-vehicle multiplier values ranging between 1-2.7 (!). Results from meta-analysis of these studies have been for example reported and used in evaluating capacity increase investments (see the case of our study of a metro line in Stockholm).
These estimates seem strangely high. They imply that passengers will rather travel twice as long if they can have a seat instead of to travelling in a densely crowded vehicle. These has severe ramifications for project appraisal – do you invest in increasing vehicle size, higher frequency or higher speed? In a choice experiment it is easy to indicate that you rather wait for the next vehicle or travel longer than to ride a busy vehicle. However, there was very scarce evidence that people actually do these trade-offs in reality. We therefore wanted to find out to what extent crowding impacts passenger route choices based on observed behavior. This is now possible thanks to large-scale smart card deployment.
See full paper here: “Crowding valuation in urban tram and bus transportation based on smart card data”
In this study, crowding valuation for urban tram and bus travelling is determined fully based on revealed preference data. Urban tram and bus crowding valuation is estimated in a European context based on a Dutch case study network. Based on the estimated discrete choice model, we conclude that crowding plays a significant role in passengers’ route choice in public transport. The average crowding multiplier of in-vehicle time equals 1.16 when all seats are occupied. For frequent travellers, this value is equal to 1.31. Our study results suggest that infrequent travellers do not incorporate expected crowding in their route choice. These values are significantly lower than those reported in past studies based on choice experiments.
The insights gained from our study can support the decision-making process of policy-makers, by quantifying the benefits of measures aiming to reduce crowding levels for example in a cost–benefit analysis framework.
On-demand public transport is expected to become an increasingly important component of public transport systems, facilitated by vehicle automation. The potential of rail-bound on-demand services has been largely overlooked. Together with Jesper Haverkmap, who did his master thesis in the Dutch Railways, we determine the capacity requirements of an envisaged automated on-demand rail-bound transit system which offers a direct non-stop service.
How many vehicles would such a system require? What are the costs associated with such a system? What are the track and station capacity required? What level of service will it deliver? What are the network saturation patterns given that vehicles can now perform route choice and choose the shortest path to passengers’ destination? What are the consequences for equity in service provision?
An optimization model is formulated for determining the optimal track and station platform capacities for an on-demand rail transit system so that passenger, infrastructure and operational costs are minimized. The macroscopic model allows for studying the underlying relations between technological, operational and demand parameters, optimal capacity settings and the obtained cost components.A series of sensitivity analyses are performed to test the consequences of a range of network structures, technological capabilities, operational settings, cost functions and demand scenarios for future automated on-demand rail-bound systems.
The model is applied to a series of numerical experiments followed by its application to part of the Dutch railway network. The performance is benchmarked against the existing service, suggesting that in-vehicle times can be reduced by 10% in the case study network while the optimal link and station capacity allocation is comparable to those currently available in the case study network. While network geometry and demand distribution are always the underlying determinants of both service frequencies and in-vehicle times, line configuration is only a determinant in the conventional system, whereas the automated on-demand rail service better caters for the prevailing demand relations, resulting in greater variations in service provision.
I am honored and delighted to be awarded with an ERC (European Research Council) Starting Grant. My project CriticalMaaS will develop and test concepts, theories and models for planning, operating and evaluating the dynamics of Mobility as a Service.
The project will run for 5 years and will be performed by a team of PhD students, post-doc researchers and in collaboration with colleagues within the Department of Transport and Planning at TU Delft and beyond.