Michail Makridis


Last Name

Makridis

First Name

Michail

ORCID

0000-0001-7462-4674

Organisational unit

08686 - Gruppe Strassenverkehrstechnik

Search Results

Publications1 - 10 of 118
  • Empirical study on the properties of adaptive cruise control systems and their impact on traffic flow and string stability
    Item type: Journal Article
    Makridis, Michail; Mattas, Konstantinos; Ciuffo, Biagio; et al. (2020)
    Transportation Research Record
    Adaptive cruise control (ACC) systems are standard equipment in many commercially available vehicles. They are considered the first step of automation, and their market penetration rate is expected to rise, along with the interest of researchers worldwide to assess their impact in relation to traffic flow and stability. These properties are currently discussed mainly through microsimulation studies and empirical observations, with the first being the most common. Experimental observations can draw safer conclusions about the behavior of such systems, but the literature is limited. In this work, an experimental campaign with five vehicles equipped with ACC was conducted at the proving ground of AstaZero in Sweden to improve understanding on the properties of ACC systems and their functionality under real driving conditions. The main parameters under investigation are the response time of controllers, the available time headway settings, and the stability of the car-platoon. The results show that the response time range for the controllers is between 1.7 and 2.5 s, significantly longer than the values reported in the literature. The range of the time headway settings was found to be quite broad. Finally, a dataset of perturbations on a variety of equilibrium speeds of the car-platoon and of variable magnitudes was created. Results clearly highlight the instability of the car-platoon. Instability is also displayed even for slight perturbations derived by variability in the road gradient. Numerical differentiation on the altitude shows a negative correlation with the speed trajectory of the leading vehicle.
  • OptFlow
    Item type: Report
    Genser, Alexander; Makridis, Michail; Kouvelas, Anastasios (2022)
    Im Zuge des Projektes «Opt-Flow» in Zusammenarbeit mit der Dienstabteilung Verkehr (DAV) der Stadt Zürich, wurde eine Beurteilung der Genauigkeit von Verkehrsfluss- und Reisezeitmessung durch verschiedene Sensorik vorgenommen. Von besonderer Interesse sind dabei neu installierte Thermalkameras mit eingebauter WiFi-Schnittstelle. Es werden Verkehrsflüsse und Reisezeiten in einem Testperimeter (in Folge auch Untersuchungsgebiet genannt) in Oerlikon, Zürich erfasst und mit diversen neuartigen Datenquellen verglichen. Final wird die Qualität der ermittelten Verkehrsgrössen in zwei Fallstudien mit Resultaten von manuell ausgewerteten Messkampagnen beurteilt. Zudem wird eine Fallstudie, basierend auf einer deskriptiven Datenanalyse, durchgeführt.
  • The impact of automation and connectivity on traffic flow and CO2 emissions
    Item type: Journal Article
    Makridis, Michail; Mattas, Konstantinos; Mogno, Caterina; et al. (2020)
    Atmospheric Environment
    The interest on the impact of vehicle automation and connectivity in the future road transport networks is very high, both from a research and a policy perspective. Results in the literature show that many of the anticipated advantages of connected and automated vehicles or automated vehicles without connectivity (CAVs and AVs respectively) on congestion and energy consumption are questionable. Some studies provide quantitative answers to the above questions through microsimulation but they systematically ignore the realistic simulation of vehicle dynamics, driver behaviour or instantaneous emissions estimates, mostly due to the overall increased complexity of the transport systems and the need for low computational demand on large-scale simulations. However, recent studies question the capability of common car-following models to produce realistic vehicle dynamics or driving behaviour, which directly impacts emissions estimations as well. This work presents a microsimulation study that contributes on the topic, using a scenario-based approach to give insights regarding the impact of CAVs and AVs on the evolution of emissions over a highway network. The motivation here is to answer whether the different driving behaviours produce significant differences in emissions during rush hours, and how significant is the impact of detailed vehicle dynamics simulation and instantaneous emissions in the outcome. The status of the network is assessed in terms of flow and speed. Furthermore, emissions are estimated using both the average-speed EMEP/EEA fuel consumption factors and a generic version of the European Commission's CO2MPAS model that provides instantaneous fuel consumption estimates. The simulation results of this work show that AVs can deteriorate the status of the network, and that connectivity is the key for improved traffic flow. Emissions-wise, the AVs have the highest fuel consumption per km travelled among other types, while CAVs only marginally lower the overall consumption of human-driven vehicles. For the same traffic demand, the total emissions for different vehicle types remain at comparable levels.
  • Network-wide freeway traffic estimation using sparse sensor data: A Dirichlet graph auto-encoder approach
    Item type: Journal Article
    Zhou, Qishen; Zhang, Yifan; Makridis, Michail; et al. (2025)
    IEEE Transactions on Intelligent Transportation Systems
    Network-wide Traffic State Estimation (TSE), which aims to infer a complete image of network traffic states with sparsely deployed sensors, plays a vital role in intelligent transportation systems. With the development of data-driven methods, traffic dynamics modeling has advanced significantly. However, TSE poses fundamental challenges for data-driven approaches, since historical patterns cannot be learned locally at sensor-free segments. Although graph representation learning shows promise in estimating states at locations without sensors, existing methods typically handle unobserved locations by filling them with zeros, introducing bias to the sensitive graph message propagation. The recently proposed Dirichlet Energy-based Feature Propagation (DEFP) method achieves State-Of-The-Art (SOTA) performance in unobserved node classification by eliminating the need for zero-filling. However, applying it to TSE faces three key challenges: inability to handle directed traffic networks, strong assumptions in traffic spatial correlation modeling, and overlooking distinct propagation rules of different patterns (e.g., congestion and free flow). We propose DGAE, a novel inductive graph representation model that addresses these challenges through theoretically derived DEFP for Directed graph (DEFP4D), enhanced spatial representation learning via DEFP4D-guided latent space encoding, and physics-guided propagation mechanisms that separately handle congested and free-flow patterns. Experiments on three traffic datasets demonstrate that DGAE outperforms existing SOTA methods and exhibits strong cross-city transferability. Furthermore, DEFP4D can serve as a standalone lightweight solution, showing superior performance under extremely sparse sensor conditions. The code of this work is publicly available at: https://github.com/ZJU-TSELab/DGAE
  • Is calibration of car-following model on spacing enough for autonomous vehicles?
    Item type: Other Conference Item
    Zheng, Shi-Teng; Makridis, Michail; Kouvelas, Anastasios (2022)
    Car-following (CF) models are used in traffic simulation scenarios aiming to provide realistic modeling of vehicles’ longitudinal movement. In most works, calibration is performed on trajectories of leading and following vehicles either on speed or spacing quantities, with the second to be considered as the most appropriate according to the literature. With the advent of adaptive cruise control (ACC) technology, complexity in simulation of vehicles’ longitudinal movement rises. There are intrinsic behavioral differences between human- and ACC-driven vehicles that are visible in experimental observations. Some examples include human-like response time, string instability, hysteretic behavior and others. Since calibration is performed only on spacing, even after proper parametrization we cannot be certain that CF models will be able to reproduce some or all of the above phenomena and to what extend. The aim of this work is to propose a multi-objective calibration framework validated on empirical observations of car platoons. Furthermore, it investigates if, and to what extent, the traffic dynamics and behavioral patterns of ACC vehicles can be reproduced. Two state-of-the-art CF models are calibrated with empirical datasets based on spacing and with the proposed framework. The capability of the models to reproduce the ACC behavior across different dimensions is discussed along with advantages and downsides of each solution.
  • Investigating link- and network-level bicycle traffic flow characteristics using a microsimulation approach
    Item type: Conference Paper
    Ni, Ying-Chuan; Makridis, Michail; Kouvelas, Anastasios (2023)
    hEART 2023: 11th Symposium of the European Association for Research in Transportation
    Implementing cycling infrastructure for road users has become a popular transport policy for cities to create a sustainable urban environment nowadays. A thorough understanding of bicycle traffic is required to evaluate new infrastructure designs. To fill the remaining knowledge gap in this aspect, this study aims to investigate bicycle traffic flow characteristics on dedicated bike lanes. A microsimulation tool is used to simulate various scenarios and compute bicycle traffic states. From the simulation results, bicycle flow characteristics presented at both link and network levels are identified and discussed. The findings are expected to be applied to future research regarding large-scale bicycle traffic flow modeling.
  • An experimental urban case study with various data sources and a model for traffic estimation
    Item type: Journal Article
    Genser, Alexander; Hautle, Noel; Makridis, Michail; et al. (2022)
    Sensors
    A reliable estimation of the traffic state in a network is essential, as it is the input of any traffic management strategy. The idea of using the same type of sensors along large networks is not feasible; as a result, data fusion from different sources for the same location should be performed. However, the problem of estimating the traffic state alongside combining input data from multiple sensors is complex for several reasons, such as variable specifications per sensor type, different noise levels, and heterogeneous data inputs. To assess sensor accuracy and propose a fusion methodology, we organized a video measurement campaign in an urban test area in Zurich, Switzerland. The work focuses on capturing traffic conditions regarding traffic flows and travel times. The video measurements are processed (a) manually for ground truth and (b) with an algorithm for license plate recognition. Additional processing of data from established thermal imaging cameras and the Google Distance Matrix allows for evaluating the various sensors’ accuracy and robustness. Finally, we propose an estimation baseline MLR (multiple linear regression) model (5% of ground truth) that is compared to a final MLR model that fuses the 5% sample with conventional loop detector and traffic signal data. The comparison results with the ground truth demonstrate the efficiency and robustness of the proposed assessment and estimation methodology.
  • Quantitative Fairness - A Framework For The Design Of Equitable Cybernetic Societies
    Item type: Working Paper
    Riehl, Kevin; Makridis, Michail; Kouvelas, Anastasios (2024)
    arXiv
    Advancements in computer science, artificial intelligence, and control systems of the recent have catalyzed the emergence of cybernetic societies, where algorithms play a significant role in decision-making processes affecting the daily life of humans in almost every aspect. Algorithmic decision-making expands into almost every industry, government processes critical infrastructure, and shapes the life-reality of people and the very fabric of social interactions and communication. Besides the great potentials to improve efficiency and reduce corruption, missspecified cybernetic systems harbor the threat to create societal inequities, systematic discrimination, and dystopic, totalitarian societies. Fairness is a crucial component in the design of cybernetic systems, to promote cooperation between selfish individuals, to achieve better outcomes at the system level, to confront public resistance, to gain trust and acceptance for rules and institutions, to perforate self-reinforcing cycles of poverty through social mobility, to incentivize motivation, contribution and satisfaction of people through inclusion, to increase social-cohesion in groups, and ultimately to improve life quality. Quantitative descriptions of fairness are crucial to reflect equity into algorithms, but only few works in the fairness literature offer such measures; the existing quantitative measures in the literature are either too application-specific, suffer from undesirable characteristics, or are not ideology-agnostic. Therefore, this work proposes a quantitative, transactional, distributive fairness framework, which enables systematic design of socially feasible decision-making systems. Moreover, it emphasizes the importance of fairness and transparency when designing algorithms for equitable, cybernetic societies.
  • Estimation of dynamic origin-destination demand for urban road networks incorporating a region-level traffic flow model
    Item type: Other Conference Item
    Ni, Ying-Chuan; Schönhärl, Anna M.; Makridis, Michail; et al. (2025)
    hEART 2025: List of Abstracts
    Representative dynamic origin-destination (OD) demand is crucial for the assessment of network traffic performance. However, challenges for OD estimation include its underdetermination nature, data scarcity, and high computation cost. This paper proposes a mathematical optimization method for dynamic OD estimation (DODE) incorporating a link-level linear mapping method and a region-level traffic flow model. The region production and accumulation are estimated from loop detector data. It is considered an efficient method particularly for a large-scale network. A case study is conducted for a real-world network. After the optimization, we further test the estimated demand with microscopic traffic simulation which has not been calibrated. It is found that approximately 50% of the link count measurements already have GEH statistic values that are below 5. More examinations are required to validate the proposed method. The outcome can also be used as the initial solution in DODE using a simulation-based optimization approach.
  • Automatisiertes Fahren im Strassentunnel
    Item type: Report
    van Linn, Andreas; Schneider, Reto; Kouvelas, Anastasios; et al. (2023)
    Selbstfahrende intelligente Fahrzeuge bringen ein neues Fahrverhalten in die Strassenverkehrssysteme ein, welches die derzeitigen Verkehrsnetze verändern wird, besonders in Anbetracht des noch über lange Zeit zu erwartenden Mischverkehrs. Die strategische Ausrichtung des ASTRA umfasst das ausdrückliche Ziel, dass bis im Jahr 2030 auf Nationalstrassen hoch- und vollautomatisierte Fahrzeuge verkehren können. Zur Zielerreichung sind sowohl die gesellschaftliche Akzeptanz dieser neuen Mobilitätsform zu fördern als auch die technische und betriebliche Machbarkeit zu testen. Eine wichtige Voraussetzung für erste Testanwendungen unter realen Verkehrsbedingungen ist die Ermittlung idealer Strecken. In diesem Forschungsprojekt haben wir folgende Hypothese untersucht: "Tunnel auf Hochleistungsstrassen sind prädestinierte Infrastrukturen zur Förderung des hoch- und vollautomatisierten Fahrens". Weil es für solche Systeme noch keine standardisierten Betriebsmerkmale und Erfahrungswerte aus der Praxis gibt, haben den aktuellen Stand der Forschung und Technik zusammengefasst. Daraus wurde ersichtlich, dass das Verhalten automatisierter Fahrzeuge unter bestimmten Infrastrukturbedingungen, wie z. B. in Tunneln und bei ungünstigen Wetterbedingungen, weiterhin unklar bleibt. Während die Sicherheit und Effizienz des Fahrens in Tunneln derzeit mit Faktoren wie Ablenkung des Fahrers, Beleuchtung und Verkehrsfluss in Verbindung gebracht werden, stellt die Einführung hochautomatisierter Fahrzeuge eine grosse Herausforderung für das Verständnis ihrer Auswirkungen auf die Tunnelsicherheit dar. Die heute für das automatisierte Fahren verwendeten Sensoren wie Kameras, LiDARs und Radare zeigen unter realen Bedingungen einen erheblichen Leistungsabfall, und es werden massgeschneiderte Testverfahren für ADAS benötigt, um den ordnungsgemässen Betrieb autonomer Fahrzeuge zu gewährleisten. Um diesen Wissenslücken zu füllen und die Hypothese zu testen, wurden Expertenworkshops durchgeführt und ein Fragebogen kreiert, mit welchem Kriterien für geeignete Tunnelstrecken definiert wurden. Mit diesen Kriterien haben wir ausgehend von der TUSI-Liste 23 Tunnels als mögliche Teststrecken für autonomes Fahren ermittelt. In einer anschliessenden Diskussion mit den ASTRA Filialen wurden Liste nochmals eingeengt und konkret 7 Tunnels als gut geeignet vorgeschlagen. Das Forschungsprojekte zeigt auf, dass die Modernisierung von Tunneln zur Unterstützung automatisierter Fahrzeuge notwendig erscheint, diese beinhaltet unter anderem geeignete Lichtverhältnisse, Unfallerkennungsverfahren, Kommunikations- infrastrukturen und klare Fahrbahnmarkierungen. AVs und CAVs haben zusätzlich das Potenzial, die Verkehrssicherheit zu verbessern und die Kapazität durch Kommunikation und Informationsaustausch zwischen Fahrzeugen und Infrastruktur zu erhöhen. Tunnel können somit auch als Informationszentren für Fahrer genutzt werden, um sie sowohl in normalen als auch in Notfallsituationen zu unterstützen. Das vorliegenden Forschungsprojekt bestätigt die Hypothese, dass «Tunnel auf Hochleistungsstrassen prädestinierte Infrastrukturen zur Förderung des hoch- und vollautomatisierten Fahrens sind. Offen bleibt die Frage, welche flankierenden Massnahmen noch getroffen werden müssen, um Tests mit automatisierten Fahrzeugen im realen Verkehrsbetrieb mit Mischverkehr durchzuführen. Daraus ergibt sich folgender Forschungsbedarf: Es gilt standardisierte Testbedingungen in Tunneln zu entwickeln. Wie müssen Tunnelwände beschaffen sein, um die Sensorik möglichst nicht zu stören? Welche Tunnelbeschaffenheit kann Reflexionen verursachen und unter welchen Umständen werden Sensoren beeinträchtigt. Um diese Unklarheiten zu beseitigen sind Tests unter realen Bedingungen notwendig. Weiter wird die Entwicklung von Testprotokollen und einer standardisierten Messmethodik zur Bewertung der Leistung von Sensoren Gegenstand zukünftiger Forschungsprojekte sein.
Publications1 - 10 of 118