Milos Balac


Last Name

Balac

First Name

Milos

ORCID

0000-0002-6099-7442

Organisational unit

02261 - Center for Sustainable Future Mobility / Center for Sustainable Future Mobility

Filters

2021 - 202627
Reset filters

Search Results

Publications 1 - 10 of 27
  • Empowering policymakers with new tools: The case of Agent-based simulation
    Item type: Other Conference Item
    Balac, Milos (2025)
  • What we learned so far with E-Bike City Research D-BAUG Lighthouse Project
    Item type: Presentation
    Elliot, Catherine; Axhausen, Kay W.; Marra, Alessio Daniele; et al. (2024)
  • Assessing the potential of feeder DRT services in urban outskirts
    Item type: Other Conference Item
    Mahfouz, Hussein; Balac, Milos; Morgan, Malcolm; et al. (2025)
  • Probabilistic routing for agent-based simulations
    Item type: Working Paper
    Meister, Adrian; Liang, Zheng; Balac, Milos (2024)
    Arbeitsberichte Verkehrs- und Raumplanung
    Traditional route assignment approaches in agent-based models often rely on least-cost path algorithms, which may not accurately reflect the complex decision-making processes of cyclists. This research addresses these limitations by incorporating probabilistic elements into the routing model, thus accommodating the variability in route choices observed in real-world scenarios. The proposed model integrates a Recursive Logit framework to account for the influence of various factors such as gradient, surface quality, traffic conditions, and dedicated cycling infrastructure on cyclists’ route selection. A case study using a detailed Zurich scenario demonstrates the model’s application and effectiveness. Results show that the probabilistic routing model not only aligns more closely with observed cyclist behavior but also offers a robust tool for urban planning and policy evaluation aimed at promoting sustainable and active transportation modes.
  • A Two-Stage Deep Learning Based Approach for Predicting Instantaneous Vehicle Speed Profiles on Road Networks
    Item type: Conference Paper
    Dib, Abdelkader; Sciarretta, Antonio; Balac, Milos (2023)
    2023 IEEE 26th International Conference on Intelligent Transportation Systems (ITSC)
    Prediction of vehicle speed profiles is vital to many transportation and vehicular applications. However, accurately predicting these profiles remains challenging due to the complex and uncertain nature of the factors influencing driver behavior. This research presents a novel data-driven model that leverages a deep two-stage long short-term memory (LSTM) architecture to effectively capture the relationship between vehicle speed and macroscopic road attributes. Our model integrates road features, obtainable from various online map services, and average speeds as inputs and generates naturalistic speed profiles for a given route. The ultimate goal of this study is to integrate the proposed model with a microscopic emission model and incorporate it downstream of a mesoscopic traffic model. This integration enables the generation of high-resolution spatiotemporal maps of traffic emissions. The proposed model is trained and evaluated on a large dataset, including various driving records. The results demonstrate its ability to accurately generate realistic driving patterns while reproducing fuel consumption and emissions levels similar to those of real-world profiles.
  • OD-Vis: A Map-based Dashboard for Insight Discovery of OD Data
    Item type: Other Conference Item
    Zuo, Chenyu; Balac, Milos; Grübel, Jascha; et al. (2024)
    Abstracts of the ICA
  • Collaborative Mobility Digital Platform
    Item type: Other Conference Item
    Balac, Milos; Reck, Daniel Jan; Bris, Myriam; et al. (2025)
    Understanding human mobility and forecasting the impacts of policy interventions or infrastructure investments are essential to transport planning and reaching societal climate goals. In Switzerland, aggregated and proprietary transport models are predominantly used for this task. This situation poses several challenges. First, these models are costly to set up and maintain. This limits their use to large organizations that can afford them, effectively excluding many interested players, such as transport operators, cities, and universities, from participating in transport planning in a more proactive way. Second, these models are inadequate to model our increasingly connected (shared) mobility systems and modern transport policies such as congestion pricing. Third, these models rarely fully utilize novel data collection techniques, such as real-time automatic passenger counts in buses and trams. As a result, societies currently underutilize their potential to solve one of the greatest challenges of our times. We propose to create an open-source collaborative platform for mobility analysis and forecasting. This platform will introduce several innovations: it will use state-of-the-art mobility data, automate data import, enable mobility data analysis, automate calibration, allow the study of complex integration of public transport with other modes, and enable policy-oriented visualization; it will use the state- of-the-art agent-based tools, eqasim and MATSim, and national ORD infrastructure, as well as the Open Digital Twin Platform, to enable robust mobility digital twin development; it will create social and economic value by bringing state-of-the-art transport modeling tools to a broader public in a cost-efficient way. The approach will be applied to Geneva, with tpg as the implementation partner. Generalizability and transferability will be demonstrated in the Lausanne agglomeration, with the Ville de Lausanne and tl as the implementation partners.
  • Route choice model integration into MATSim
    Item type: Other Conference Item
    Meister, Adrian; Balac, Milos; Zheng, Liang; et al. (2024)
    This paper presents the integration of explicit discrete route choice models into the agentbased simulation framework MATSim, as an example. It represents an obvious research direction, which to the best of the authors’ knowledge has not yet been presented for any other agent-based transport simulation framework. Discrete route choice models, estimated from stated- or revealed preference data, are backed by years of research and can be effectively used for prediction. They allow to realistically model heterogeneity using econometric theory, and typically allow for faster model convergence towards user equilibria. We describe the technical integration of such models into MATSim and demonstrate the results using a scenario of Zurich. In a first step, we implement the route choice model only for cycling, but stress that our method is applicable to any non-PT mode.
  • Simulation of policies for automated ride-hailing and ride-pooling services
    Item type: Working Paper
    Kagho, Grace Orowo; Balac, Milos; van Eggermond, Michael A.B.; et al. (2024)
    Arbeitsberichte Verkehrs- und Raumplanung
    Automated vehicles are becoming more prevalent, and the disruption they would cause in combination with ride-hailing and ride-pooling services could be tremendous. Therefore, this study investigates the impacts of ride-hailing and ride-pooling automated fleets in two Swiss cities, Chur and Zurich, and potential policy measures to steer their operations towards more sustainable solutions. We employ the results of the stated preference survey and combine the estimated mode-choice and car ownership model results with the agent-based simulation, MATSim, to simulate the impacts of various scenarios. We find that automated ride-hailing (aRH) and automated ride-pooling (aRP) services do not seem to be competing for the same demand. In general, these services would lead to a reduction in total travel time but an increase in total vehicle distance, which is more substantial in transit-oriented Zurich than in car-oriented Chur. Furthermore, we found that even though the proposed policies increased vehicle occupancy, they did not manage to overcome the increase in VKT, signaling the need for more targeted policies and operational strategies. Finally, we provide recommendations for transport policy and future research based on our findings.
  • Multimodality in the Swiss New Normal (SNN)
    Item type: Other Conference Item
    Heimgartner, Daniel; Sallard, Aurore; Balac, Milos; et al. (2024)
Publications 1 - 10 of 27