Torsten Bertram

Technische Universität Dortmund | TUD · Robotics Research Institute

Distribution shifts between operational domains can severely affect the performance of learned models in self-driving vehicles (SDVs). While this is a well-established problem, prior work has mostly explored naive solutions such as fine-tuning, focusing on the motion prediction task. In this work, we explore novel adaptation strategies for differen...

Accurate 3D object detection is vital for automated driving. While lidar sensors are well suited for this task, they are expensive and have limitations in adverse weather conditions. 3+1D imaging radar sensors offer a cost-effective, robust alternative but face challenges due to their low resolution and high measurement noise. Existing 3+1D imaging...

Effective tracking of surrounding traffic participants allows for an accurate state estimation as a necessary ingredient for prediction of future behavior and therefore adequate planning of the ego vehicle trajectory. One approach for detecting and tracking surrounding traffic participants is the combination of a learning based object detector with...

In driver monitoring various data types are collected from drivers and used for interpreting, modeling, and predicting driver behavior, and designing interactions. Aim of this contribution is to introduce manD 1.0, a multimodal dataset that can be used as a benchmark for driver monitoring in the context of automated driving. manD is the short form...

The workshop proceedings contain the contributions of the 33rd workshop "Computational Intelligence" which will take place from 23.11. - 24.11.2023 in Berlin. The focus is on methods, applications and tools for ° Fuzzy systems, ° Artificial Neural Networks, ° Evolutionary algorithms and ° Data mining methods as well as the comparison of methods on...

This work uses game theory as a mathematical framework to address interaction modeling in multi-agent motion forecasting and control. Despite its interpretability, applying game theory to real-world robotics, like automated driving, faces challenges such as unknown game parameters. To tackle these, we establish a connection between differential gam...

Game theory offers an interpretable mathematical framework for modeling multi-agent interactions. However, its applicability in real-world robotics applications is hindered by several challenges, such as unknown agents' preferences and goals. To address these challenges, we show a connection between differential games, optimal control, and energy-b...

New 3+1D high-resolution radar sensors are gaining importance for 3D object detection in the automotive domain due to their relative affordability and improved detection compared to classic low-resolution radar sensors. One limitation of high-resolution radar sensors, compared to lidar sensors, is the sparsity of the generated point cloud. This spa...

Recent developments and the beginning market introduction of high-resolution imaging 4D (3+1D) radar sensors have initialized deep learning-based radar perception research. We investigate deep learning-based models operating on radar point clouds for 3D object detection. 3D object detection on lidar point cloud data is a mature area of 3D vision. M...

The Steer-by-Wire (SbW) system is a key technology for highly automated driving. For automated lateral vehicle guidance, the precise position control of the SbW Front Axle Actuator is an essential prerequisite. This contribution presents the modeling, control design, nominal performance, and stability analysis as well as the robustness analysis of...

Offline reinforcement learning (RL) provides a framework for learning decision-making from offline data and therefore constitutes a promising approach for real-world applications such as automated driving (AD). Especially in safety-critical applications, interpretability and transferability are crucial to success. That motivates model-based offline...

There are significant advances in GNSS-free cross-modality self-localization of self-driving vehicles. Recent methods focus on learnable features for both cross-modal global localization via place recognition (PR) and local pose tracking, however they lack means of combining them in a complete localization pipeline. That is, a pose retrieved from P...

This paper deals with the integration of input move-blocking into the framework of suboptimal model predictive control. The blocked input parameterization is explicitly considered as a source of suboptimality. A straightforward integration approach is to hold back a manually generated stabilizing fallback solution in some buffer for the case that t...

Motion planning and control are crucial components of robotics applications. Here, spatio-temporal hard constraints like system dynamics and safety boundaries (e.g., obstacles in automated driving) restrict the robot's motions. Direct methods from optimal control solve a constrained optimization problem. However, in many applications finding a prop...

It is widely assumed that considering vehicle interactions for trajectory prediction can significantly improve accuracy. All environmental sensors of an automated vehicle (AV) suffer from occlusion. Therefore, relevant vehicles can be occluded by others, especially in dense traffic situations, and remain invisible to the AV. Unobserved vehicles cou...

The contribution at hand presents a novel approach to generate dynamic energy-optimized illumination by matrix headlamps designed for automated driving. The approach consists of a novel basic minimal illumination adapted to the current road environment with minimum brightness and a light pixel control strategy to adjust the illumination of potentia...

Zusammenfassung Eine der zentralen Problemstellungen beim bedingt- und hochautomatisierten Fahren liegt in der Gestaltung einer sicheren und komfortablen Aufgabenübertragung zwischen dem automatisierten System und dem menschlichen Fahrer und vice versa. Dieser Beitrag stellt ein holistisches Modell zur Übergabe und Übernahme von Fahraufgaben vor, w...

Zusammenfassung Um die Unfallzahlen weiter zu senken, schreibt die Europäische Union ab 2030 eine höhere Fahrerüberwachung für neue Fahrzeuge vor. Bislang liegt der Fokus in einem manuell gefahrenen Fahrzeug auf einer Müdigkeitserkennung als Komfortsystem. Jedoch ändern sich die Anforderungen an den Fahrer und dessen Aufgaben bei steigender Automat...

In automated vehicles, the collaboration of human drivers and automated systems plays a decisive role in road safety, driver comfort, and acceptance of automated vehicles. A successful interaction requires a precise interpretation and investigation of all influencing factors such as driver state, system state, and surroundings (e.g., traffic, weath...

Towards the aim of mastering level 5, a fully automated vehicle needs to be equipped with sensors for a 360∘ surround perception of the environment. In addition to this, it is required to anticipate plausible evolutions of the traffic scene such that it is possible to act in time, not just to react in case of emergencies. This way, a safe and smoot...

Trajectory optimization is a promising method for planning trajectories of robotic manipulators. With the increasing success of collaborative robots in dynamic environments, the demand for online planning methods grows and offers new opportunities as well as challenges for trajectory optimization. Special requirements in terms of real-time capabili...

Inhalt Vorwort ..... 1 Car2X-Kommunikation Bildgebende Car2Car-Kommunikation mit sichtbarem Licht . . . . . 3 Lichtbasierte Kommunikationsschnittstelle zwischen automatisierten Fahrzeugen und anderen Verkehrsteilnehmern . . . . . . . . . . . . 17 Hochauflösende Scheinwerfersysteme Verbesserung der kameragestützten Objekterkennung im Straßenverkehr...

One of the key aspects for an efficient cooperation between human driver and automated vehicle lies in the accurate interpretation of the driver state by the automated system. Flawless driver monitoring and consequently successful driver-vehicle interaction can increase safety of the traffic in the future when automated agents are one of the involv...

Motion planning and control are crucial components for automated vehicles. Especially in parking scenarios, high precision is required due to the small distances to obstacles. Common approaches utilize ultrasonic or camera sensors. This paper presents a radar-based system architecture for automated parking, which can operate more robustly under dif...

Offline reinforcement learning (RL) provides a framework for learning decision-making from offline data and therefore constitutes a promising approach for real-world applications as automated driving. Self-driving vehicles (SDV) learn a policy, which potentially even outperforms the behavior in the sub-optimal data set. Especially in safety-critica...

This paper deals with the integration of input move-blocking into the framework of suboptimal model predictive control. The blocked input parameterization is explicitly considered as a source of suboptimality. A straightforward integration approach is to hold back a manually generated stabilizing fallback solution in some buffer for the case that t...

In the contribution, a model predictive trajectory tracking approach is presented. Due to the utilization of an accurate prediction model, which considers not only the vehicle dynamics but also the limited actuator dynamics, the approach can be used even in emergency collision avoidance systems. The approach explicitly predicts a trajectory set for...

This paper evaluates different deep learning based depth estimation algorithms. We propose improvements for a state-of-the-art unguided depth completion method where the number of necessary parameters can be more than halved at unvarying accuracy. Based on the results of the depth estimation evaluation, we consider the performance of semantic segme...

Zusammenfassung In diesem Beitrag wird eine suboptimale modellprädiktive Regelung mit stabilisierenden Endbedingungen und einem ableitungsfreien Optimierungsalgorithmus für eine bestimmte Klasse nichtlinearer Systeme vorgestellt. Repräsentativ für diese Systemklasse stehen insbesondere elektromagnetische Aktuatoren. Durch die Kombination eines Frei...

This contribution presents a novel probabilistic approach for the generation of discretionary lane change proposals with a focus on highway driving situations. The developed model is based on the quantification of the utility of driving lanes. It generates a lane change proposal if the current driving lane is unsatisfactory in the sense that the de...

This paper deals with the development and analysis of novel time-optimal point-to-point model predictive control concepts for nonlinear systems. Recent approaches in the literature apply a time transformation, however, which do not maintain recursive feasibility for a piecewise constant control parameterization. The key idea in this paper is to int...

Elastic manipulators often result from lightweight construction or safety requirements in human–robot-collaboration scenarios. In many cases, vibration-damping becomes necessary to enable stable and precise operation, which imposes high demands on controllers. A fundamental challenge for link-elastic manipulators with general kinematics and no dedi...

Compared to standard navigation maps, HD maps contain precise additional information for automated vehicles. To exploit this information, a lane-level accurate localization estimate within the HD map needs to be available. Here, the computation and memory overhead of the localization algorithm needs to be as small as possible to enable a reasonable...

Die inhaltlichen Schwerpunkte des Tagungsbands zur ATZlive-Veranstaltung "Automatisiertes Fahren" zeigen die Treiber des automatisierten Fahrens auf: Künstliche Intelligenz, Machine- oder Deep-Learning. Das Zusammenspiel von künstlicher und menschlicher Intelligenz sowie die Fähigkeit von Mensch und Maschine zu kooperieren müssen in neuen Interakti...

Triggering emotions in a driving simulator is not easy as the virtual environment reduces the reality of the situations. This contribution deals with the induction of emotions in drivers during the simulation and addresses the possible hindrances in the design and implementation phases. For this purpose, an experiment is conducted on a driving simu...

This contribution presents a novel algorithm for real-time simulation of adaptive matrix- and pixel-headlights for motor vehicles. The simulation can generate the light distribution of a pair of pixel-headlamps with a resolution of more than one and a half million matrix-elements per light module in real-time. This performance is achieved by dividi...

Improving driver performance in critical driving situations is the key to increased road safety. In SAE Level 3 of automation, the driver is retained as a fallback in critical situations. Once the system reaches its limits, the distracted driver must overtake the driving task and is responsible for the present and upcoming situations. This contribu...

Environment modeling utilizing sensor data fusion and object tracking is crucial for safe automated driving. In recent years, the classical occupancy grid map approach, which assumes a static environment, has been extended to dynamic occupancy grid maps, which maintain the possibility of a low-level data fusion while also estimating the position an...

This paper proposes a sampling-based model predictive control scheme with a single degree of freedom in control. A variable horizon and stabilizing terminal conditions ensure recursive feasibility, asymptotic stability, and improvement of the closed-loop performance. The initial derivation leads to a computationally demanding mixed-integer nonlinea...

Driver assistance systems have been in use for a decade and the automated vehicles are expected to hit the market soon. Collaboration between drivers and assistance systems, especially in SAE Level 3 of driving automation, plays a significant role as it is directly related to driving safety and the acceptance of automated vehicles. This contributio...

This paper proposes a novel online motion planning approach to robot navigation based on nonlinear model predictive control. Common approaches rely on pure Euclidean optimization parameters. In robot navigation, however, state spaces often include rotational components which span over non-Euclidean rotation groups. The proposed approach applies non...

This paper deals with time-optimal control of nonlinear continuous-time systems based on direct collocation. The underlying discretization grid is variable in time, as the time intervals are subject to optimization. This technique differs from approaches that are usually based on a time transformation. Hermite-Simpson collocation is selected as com...

This paper deals with the development and analysis of novel time-optimal point-to-point model predictive control concepts for nonlinear systems. Recent approaches in the literature apply a time transformation, however, which do not maintain recursive feasibility for piecewise constant control parameterization. The key idea in this paper is to intro...

Collaborative robots have to adapt its motion plan to a dynamic environment and variation of task constraints. Currently, they detect collisions and interrupt or postpone their motion plan to prevent harm to humans or objects. The more advanced strategy proposed in this article uses online trajectory optimization to anticipate potential collisions,...

Die inhaltlichen Schwerpunkte des Tagungsbands zur ATZlive-Veranstaltung "Automatisiertes Fahren" zeigen die Treiber des automatisierten Fahrens auf: Künstliche Intelligenz, Machine- oder Deep-Learning. Das Zusammenspiel von künstlicher und menschlicher Intelligenz sowie die Fähigkeit von Mensch und Maschine zu kooperieren müssen in neuen Interakti...

The field of human-robot interaction is a typical application of elastic robots, as they reduce the risk of injuries and physical damage in case of a collision. Elasticities, however, also impose high demands on underlying joint controllers to guarantee minimal vibration during regular operation. Numerous control concepts assume a sufficiently high...

Zusammenfassung Der vorliegende Beitrag untersucht die Erkennung benachbarter Fahrstreifen auf Grundlage von Kamerabildern. Hierbei wird sowohl die Anzahl befahrbarer Fahrstreifen als auch deren Verlauf innerhalb eines festgelegten Bereichs vor dem Fahrzeug bildbasiert geschätzt. Die Erkennung erfolgt durch Convolutional Neural Networks. Der Beitra...

The article presents an overview of the status quo in benchmarking in classification and nonlinear regression. It outlines guidelines for a comparative analysis in machine learning, benchmarking principles, accuracy estimation, and model validation. It provides references to established repositories and competitions and discusses the objectives and...

Zusammenfassung Menschliche Fahrfehler stellen die Hauptursache für Unfälle im Straßenverkehr dar. Die automatische Verkehrsüberwachung bietet einen Beitrag, um die Vision des unfallfreien Straßenverkehrs zu erreichen. Eine solche Infrastruktur erhöht unmittelbar die Verkehrssicherheit insbesondere vor dem Hintergrund einer langwierigen Durchdringu...

To guide an automated vehicle safely through complex traffic, knowledge about the future evolution of the driving situation has to be considered. The contribution at hand proposes an approach for automated driving in structured environments. An environment representation for trajectory planning is presented that enables predictive driving by interc...

This contribution presents A-D-PolyC (Automated Driving using Polygon Clipping), a novel framework for lane change behavior planning of automated vehicles on highways. It assumes that a mission planning layer generates lane change requests. In crowded traffic scenes, various variants for the lane change execution arise. The developed algorithm iden...

This tutorial chapter provides a comprehensive step-by-step guide on the setup of the navigation stack and the teb_local_planner package for mobile robot navigation in dynamic environments. The teb_local_planner explicitly considers dynamic obstacles and their predicted motions to plan an optimal collision-free trajectory. The chapter introduces a...

A major challenge in automated driving is to understand what