Conference Papers 2022

Seminar
Day 2 (8 Sep 2022), Session 4, POST-COVID DEMAND PROFILES: INTERNATIONAL REVIEW, 10:00 - 11:30

Status
Accepted, documents submitted

Submitted by / Abstract owner
Joschka Bischoff

Authors
Patrick Manser
Lukas Sieber
Joschka Bischoff
Ihab Kaddoura
Patrick Bützberger

Short abstract
Predicting the impact of post-pandemic teleworking on rail demand in Switzerland

Abstract
Planning for future rail infrastructure, efficient service concepts, as well as rolling stock considering all possible changes (e.g., in population, policy, or technology) relies on quantitative forecasts of travel demand. For this reason, SBB (Swiss Federal Railways) applies the agent-based transport model called SIMBA MOBi. This model is one of the few agent-based models in Europe being applied in practice. The model supports company-wide investment decisions (short-, mid-, and long-term). The model architecture and general application portfolio was presented at the ETC 2019 and 2020.
In this contribution, we show an extension of the model to predict teleworking behaviour in the post-pandemic world. The COVID-19 pandemic has caused a major disruption in mobility behaviour and poses a key challenge in predicting future rail demand. SBB expects a substantial mid- to long-term impact (2025 onwards) particularly on commuter flows. The main driver of this change in commuter mobility is assumed to be teleworking, which has become increasingly popular and accepted during the pandemic. This means that the pandemic is expected to be a long-term catalyst for a continuing trend towards more flexible working arrangements.
We present an approach to integrate a behavioural model for teleworking into our simulation framework SIMBA MOBi. This is achieved by making usage of the agent-based model characteristics. For each agent, a work-from-home model computes the overall probability for working from home (i.e., the total share of people teleworking on an average day) as well as the relative probability for an individual agent (i.e., how much more likely is an agent to telework based on his personal attributes). The overall probability is a model assumption which has been derived from an extensive literature research and an internal survey among stakeholders.
The crucial part is the selection of agents to work from home on the simulated day based on individual decisions. These decisions have a big impact on rail demand. For example, it might be more probable to work from home for office-persons with long commutes (typically persons with an affinity for railways) compared to a hairdresser working in the same village (less rail-oriented). We derive individual probabilities from the nationwide household survey “Microcensus Mobility and Transport”. In this survey, conducted in 2015, a part of the sample was asked about their options regarding teleworking. Using a pre-pandemic survey to predict post-pandemic impacts comes with an assumption: the relative probability for telecommuting between different types of agents is not affected by the pandemic, but we scale the overall probability for working from home to levels.
At the conference, we demonstrate our likelihood estimation routine and discuss the model attributes that are most relevant for the relative probability for teleworking (person attributes as well as characteristics of home and workplace). Then, we introduce a procedure to scale the overall probability for teleworking. The final probability set is then applied to the synthetic population in SIMBA MOBi and the individual decisions about teleworking are simulated. Based on these decisions, new daily mobility plans are generated and simulated in a dynamic traffic assignment using the software MATSim. We conclude with a detailed result analysis of the individuals working from home as well as of the overall rail demand compared to the baseline scenario. Finally, we discuss further steps to extend the model. One of the key features which will be addressed in the future are changes in mobility behaviour that are induced by teleworking, in particular the increase in leisure trips to compensate for reduced commute trips.

Programme committee
Transport Models

Topic
Behaviour Change – the impacts of the climate emergency and COVID-19 on long term travel patterns