Jens Kober

Collision avoidance – involving a rapid threat detection and quick execution of the appropriate evasive maneuver – is a critical aspect of driving. However, existing models of human collision avoidance behavior are fragmented, focusing on specific scenarios or only describing certain aspects of the avoidance behavior, such as response times. This paper addresses these gaps by proposing a computational cognitive model of human collision avoidance behavior based on active inference. Active inference provides a unified approach to modeling human behavior: the minimization of free energy. Building on prior active inference work, our model incorporates established cognitive mechanisms such as evidence accumulation to simulate human responses in three distinct collision avoidance scenarios: front-to-rear lead vehicle braking, lateral incursion by an oncoming vehicle, and another vehicle failing to yield at an intersection. We demonstrate that our model explains a wide range of empirical findings on human collision avoidance behavior. Specifically, the model closely reproduces both aggregate results from meta-analyses previously reported in the literature and detailed, scenario-specific effects observed in two recent driving simulator studies, including response timing, maneuver selection, and execution. Our results highlight the potential of active inference as a generalizable framework for understanding and modeling human behavior in complex real-life driving tasks.

To reliably perform everyday tasks, collaborative robots must be accurate, not merely repeatable. Unfortunately, precise kinematic calibration often relies on tools that are more expensive than the robots themselves. We address this limitation by proposing a low-cost and effective calibration method aimed at democratizing cobot calibration. Our minimalist approach uses a single 3D-printable two-socket spherical-joint tool to kinematically constrain the robot end effector during data collection. An optimization routine updates the nominal kinematic model to ensure consistent socket predictions while preserving their mean distance. We validate the method on Franka, KUKA, and Kinova cobots, consistently reducing mean absolute errors, for example, from approximately 10 mm to 0.2 mm on Franka robots. To demonstrate practical impact, we further evaluate the calibrated model on a Franka robot in a peg-in-the-hole task with 0.4 mm tolerance and in a repeated drawing task using Cartesian control and learning from demonstration. Both tasks fail without calibration and consistently succeed with the calibrated model. The proposed method enables affordable and practical cobot calibration, providing a foundation for accurate manipulation tasks.

Robot learning from humans has been proposed and researched for several decades as a means to enable robots to learn new skills or adapt existing ones to new situations. Recent advances in AI, including learning approaches like reinforcement learning and architectures like transformers and foundation models, combined with access to massive datasets, have created attractive opportunities to apply those data-hungry techniques to this problem. We argue that the focus on massive amounts of pre-collected data, and the resulting learning paradigm, where humans demonstrate and robots learn in isolation, is overshadowing a specialized area of work we term Human-Interactive Robot Learning (HIRL). This paradigm, wherein robots and humans interact during the learning process , is at the intersection of multiple fields (AI, robotics, human–computer interaction, design and others) and holds unique promise. Using HIRL, robots can achieve greater sample efficiency (as humans can provide task knowledge through interaction), align with human preferences (as humans can guide the robot behavior toward their expectations), and explore more meaningfully and safely (as humans can utilize domain knowledge to guide learning and prevent catastrophic failures). This can result in robotic systems that can more quickly and easily adapt to new tasks in human environments. The objective of this article is to provide a broad and consistent overview of HIRL research and to guide researchers toward understanding the scope of HIRL, and current open or underexplored challenges related to four themes—namely, human, robot learning, interaction, and broader context. The article includes concrete use cases to illustrate the interaction between these challenges and inspire further research according to broad recommendations and a call for action for the growing HIRL community.

Learning from demonstration (LfD) has proved useful for teaching robots complex skills with high sample efficiency. However, teaching long-horizon tasks with multiple skills is challenging as deviations tend to accumulate, the distributional shift becomes more evident, and human teachers become fatigued over time, thereby increasing the likelihood of failure. To address these challenges, we introduce (ST)² a sequential method for learning long-horizon manipulation tasks that allows users to control teaching flow by specifying keypoints, enabling structured and incremental demonstrations. Using this framework, we study how users respond to two teaching paradigms: 1) a traditional monolithic approach in which users demonstrate the entire task trajectory at once, and 2) a sequential approach, in which the task is segmented and demonstrated step by step. We conducted an extensive user study on the restocking task with 16 participants in a realistic retail store environment, evaluating the user preferences and effectiveness of the methods. A user-level analysis showed superior performance for the sequential approach in most cases (10 users), compared with the monolithic approach (five users), with one tie. Our subjective results indicate that some teachers prefer sequential teaching—as it allows them to teach complicated tasks iteratively—or others prefer teaching in one go due to its simplicity.

Tactile sensing is a fundamental enabler of robust interaction in environments characterized by dense and uncertain contacts. Yet, existing tactile sensors remain far from human inspiration: they are often mechanically rigid or restricted in the richness of the signals they provide. This paper aims to make a step towards solving this challenge by introducing TacLight: a compact, camera-free optical tactile sensor capable of detecting all six independent force/torque components—normal, shear, torsional, and twisting forces. The sensor is also capable of localizing the contact point. The sensing principle relies on infrared LEDs that emit light through a transparent porous elastomer, while a thin white reflective layer on top enhances light reflection, with distributed photoresistors registering transmitted intensities. Deformations of the porous medium perturb internal light paths, producing direct variations in the electrical output that encode contact mechanics without computationally heavy reconstruction. TacLight achieves high sensitivity (100 mV/N in the 0–3 N range), maintains excellent linearity (R2 > 0.95) with a 10 mm elastomer layer, and preserves a compact, low-cost, and scalable architecture. Two demonstrations highlight TacLight’s versatility. In human–machine interaction, a joystick-like device enables intuitive drone control in a simulation environment. In robotic manipulation, tactile feedback improves the success rate of an imitation-learning-based peeling task compared with a baseline without tactile sensing.

Deep generative models such as conditional variational autoencoders (CVAEs) have shown great promise for predicting trajectories of surrounding agents in autonomous vehicle planning. State-of-the-art models have achieved remarkable accuracy in such prediction tasks. Besides accuracy, diversity is also crucial for safe planning because human behaviors are inherently uncertain and multimodal. However, existing methods generally lack a scheme to generate controllably diverse trajectories, which is arguably more useful than randomly diversified trajectories, to the end of safe planning. To address this, we propose PrefCVAE, an augmented CVAE framework that uses weakly labeled preference pairs to imbue latent variables with semantic attributes. Using average velocity as an example attribute, we demonstrate that PrefCVAE enables controllable, semantically meaningful predictions without degrading baseline accuracy. Our results show the effectiveness of preference supervision as a cost-effective way to enhance sampling-based generative models.

While trajectory prediction plays a critical role in enabling safe and effective path-planning in automated vehicles, standardized practices for evaluating such models remain underdeveloped. Recent efforts have aimed to unify dataset formats and model interfaces for easier comparisons, yet existing frameworks often fall short in supporting heterogeneous traffic scenarios, joint prediction models, or user documentation. In this work, we introduce STEP -- a new benchmarking framework that addresses these limitations by providing a unified interface for multiple datasets, enforcing consistent training and evaluation conditions, and supporting a wide range of prediction models. We demonstrate the capabilities of STEP in a number of experiments which reveal 1) the limitations of widely-used testing procedures, 2) the importance of joint modeling of agents for better predictions of interactions, and 3) the vulnerability of current state-of-the-art models against both distribution shifts and targeted attacks by adversarial agents. With STEP, we aim to shift the focus from the ``leaderboard'' approach to deeper insights about model behavior and generalization in complex multi-agent settings.

This paper presents an Impedance Primitive-augmented hierarchical reinforcement learning framework for efficient robotic manipulation in sequential contact tasks. We leverage this hierarchical structure to sequentially execute behavior primitives with variable stiffness control capabilities for contact tasks. Our proposed approach relies on three key components: an action space enabling variable stiffness control, an adaptive stiffness controller for dynamic stiffness adjustments during primitive execution, and affordance coupling for efficient exploration while encouraging compliance. Through comprehensive training and evaluation, our framework learns efficient stiffness control capabilities and demonstrates improvements in learning efficiency, compositionality in primitive selection, and success rates compared to the state-of-the-art. The training environments include block lifting, door opening, object pushing, and surface cleaning. Real world evaluations further confirm the framework's sim2real capability. This work lays the foundation for more adaptive and versatile robotic manipulation systems, with potential applications in more complex contact-based tasks.