Nico Steinhardt, Raphael Wenzel, Malte Probst , Markus Amann, "Lateral Model Predictive Control for Autonomous Vehicle Prototypes", IFAC World Congress 2023, 2023.
AbstractThis paper shows a (lateral) Model Predictive Control (MPC) implementation on an Autonomous Driving (AD) prototype. Rapid prototyping and testing of AD functions in a realistic environment is a crucial step to understanding the advantages and shortcomings of algorithms in research and development of AD. Prototype vehicles show a specific set of requirements which differ from the control deployed in the final products. Vehicles are equipped with s...
Radu Stoican, Angelo Cangelosi, Thomas Weisswange, "MEWA: A Benchmark For Meta-Learning in Collaborative Working Agents", IEEE Symposium Series on Computational Intelligence (SSCI 2023), 2023.
AbstractMeta-reinforcement learning aims to overcome important limitations in reinforcement learning, like low sample efficiency and poor generalization, by creating agents that adapt to new tasks. The development of intelligent robots would benefit from such agents. Long-standing issues like data collection and generalization to real-world dynamic environments could be mitigated by sample-efficient adaptable algorithms. However, most such algorithms ...
Frank Joublin and Antonello Ceravola, "CoPAL: Corrective Planning of Robot Actions with Large Language Models", Artificial Intelligence Meetup Frankfurt, 2023.
AbstractIntroduction of HRI-EU at AI Meetup Frankfurt and presentation of the research done on a robotic system using Large Language Models (LLMs) for task and motion planning. The architecture combines reasoning, planning, and motion, with a special focus on correcting plan errors. Its efficiency is tested in simulations and real-world tasks for tasks like block arrangement, cocktail and pizza preparation....
Christiane Wiebel, "Investigating Human-Machine Cooperation", Colloquium at the Center for Cognitive Science (TU Darmstadt), 2023.
AbstractIntelligent systems have become prevalent in everyday life and keep developing at a high pace. Many of these systems do not act autonomously but operate in interaction with human users. Recent HMI research has hypothesized that such an interaction between human and machine is best reached by designing the system to behave cooperatively towards the human user (Bengler, 2012, Bütepage, 2017, Krüger, 2018, Sendhoff, 2020). In this talk, I will int...
Tim Puphal, Ryohei Hirano, Raphael Wenzel, Malte Probst , Akihito Kimata, "Considering Human Factors in Risk Maps for Robust and Foresighted Driver Warning", IEEE International Symposium on Robot and Human Interactive Communication, 2023.
AbstractDriver support systems that include human states in the support process is an active research field. Many recent approaches allow, for example, to sense the driver’s drowsiness or awareness of the driving situation. However, so far, this rich information has not been utilized much for improving the effectiveness of support systems. In this paper, we therefore propose a warning system that uses human states in the form of driver errors and can w...
Sandra Ittner, Dominik Mühlbacher, Mark Vollrath, Thomas Weisswange, "Co-DAS: Developing a Co-Driver Assistance System to Reduce Passenger Discomfort ", Frontiers in Psychology, vol. 14, pp. 17, 2023.
AbstractThe front seat passenger is often neglected when developing support systems for cars. There exist few examples of systems that provide information or interaction possibilities specifically to those passengers. Previous research indicated that the passive role of the passenger can frequently lead to a feeling of discomfort, potentially caused by missing information and missing control with respect to the driving situation. This paper proposes a va...
Stefan Menzel, Yew Ong, Yaochu Jin, Bernhard Sendhoff, "Special Session on Generative AI and Heuristic Optimization", IEEE Congress on Evolutionary Computation, 2023.
AbstractGenerative AI and Large Language Models as groundbreaking technological innovations promise to redefine the boundaries of contemporary science and engineering. Their capabilities to produce e.g., context-sensitive text, knowledge-based answers, software code, images, music and 3D assets from text prompts and image inputs create opportunities for a manifold of disciplines. Through the training on large data, these models conserve knowledge, identi...
Matti Krüger, Vanessa Krüger, Taisuke Mukai, "Evaluation of Interfaces for Augmenting a Driver's Ability to Anticipate Front Risks in Real Traffic", 27th International Technical Conference on the Enhanced Safety of Vehicles (ESV), 2023.
AbstractEffective alerts are often subject to a tradeoff between relevance and utility. While it is easier to acknowledge the relevance of a warning about an imminent hazard than a more distant threat, the possibilities to act appropriately in response to notifications decrease with threat distance. To benefit from the advantages of early notifications without creating annoyance and ignorance, we introduce a variety of Human-Machine Interfaces that pr...
Takahiro Matsuoka, Tsuyoshi Nojiri, Vanessa Krüger, Matti Krüger, "Novel interfaces that enhance a driver’s ability to perceive forward collision risks", 27th International Technical Conference on the Enhanced Safety of Vehicles (ESV), 2023.
AbstractForward Collision Warning (FCW) systems that alert a driver about the risk of rear-end collisions can contribute to a reduction of traffic accidents caused by human errors. Typically, FCWs create alerts that appear late when the risk is already high and are of binary nature, i.e., either in an alerting state during high risk or not producing any alert at lower risks. The choice at what risk level to start alerting in a binary manner is subject to...
Antonello Ceravola and Frank Joublin, "Exploring AI Architectures in the Age of LLM", Generative AI Europe 2023, 2023.
AbstractIn this talk we present at the Generative AI conference the HRI-EU institute at first, then we recap the evolution of AI in the trends of LLM and their applicability in different domains and products. We touch on the main exposed limitation of LLM and a sample of the different solution the community and the different AI companies came to. We then pick 3 investigated use-cases HRI-EU did on the usage of generative AI: Text to 3D generation in car ...