Author name cluster

Felix Huber

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

8 papers

1 author row

ICML Conference 2022 Conference Paper

Training Characteristic Functions with Reinforcement Learning: XAI-methods play Connect Four

Stephan Wäldchen
Sebastian Pokutta
Felix Huber

Characteristic functions (from cooperative game theory) are able to evaluate partial inputs and form the basis for attribution methods like Shapley values. These attribution methods allow us to measure how important each input component is for the function output—one of the goals of explainable AI (XAI). Given a standard classifier function, it is unclear how partial input should be realised. Instead, most XAI-methods for black-box classifiers like neural networks consider counterfactual inputs that generally lie off-manifold, which makes them hard to evaluate and easy to manipulate. We propose a setup to directly train characteristic functions in the form of neural networks to play simple two-player games. We apply this to the game of Connect Four by randomly hiding colour information from our agents during training. This has three advantages for comparing XAI-methods: It alleviates the ambiguity about how to realise partial input, makes off-manifold evaluation unnecessary and allows us to compare the methods by letting them play against each other.

Details

IROS Conference 2016 Conference Paper

High accuracy visual servoing for aerial manipulation using a 7 degrees of freedom industrial manipulator

Maximilian Laiacker
Felix Huber
Konstantin Kondak

This paper is devoted to the performance optimization of an aerial manipulation system composed of a Flettner-helicopter and 7 DoF manipulator. With experiments we demonstrate that the time delays in signal propagation between perception and actuation modules play an important role for the overall performance of an aerial manipulator system using visual servoing. We present an approach for estimation of the perception-action time delay and its active compensation based on the predicted motion of the manipulator end-effector.

Details

ICRA Conference 2014 Conference Paper

Aerial manipulation robot composed of an autonomous helicopter and a 7 degrees of freedom industrial manipulator

Konstantin Kondak
Felix Huber
Marc Schwarzbach
Maximilian Laiacker
Dominik Sommer
Manuel Béjar
Aníbal Ollero

This paper is devoted to a system for aerial manipulation, composed of a helicopter and an industrial manipulator. The usage of an industrial manipulator is motivated by practical applications which were identified in different cooperation projects with the industry. We address the coupling between manipulator and helicopter and show that even in case when we have an ideal controller for manipulator and a highperformance controller for helicopter, an unbounded energy flow can be generated by internal forces between helicopter and manipulator if both controllers are used independently. To solve this problem we propose a new kinematical coupling for control by introducing an additional manipulation DoF realized by helicopter rotation around its yaw axis. The new experimental setup and required modifications in the manipulator controller for this purpose are described. Further, we propose dynamical coupling which is implemented by modification of the helicopter controller feeding the interaction force/torque, measured between manipulator base and fuselage, directly to the actuators of the rotor blades. At the end, we present experimental results for aerial manipulation and their analysis.

Details

IROS Conference 2013 Conference Paper

Dynamic optimality in real-time: A learning framework for near-optimal robot motions

Roman Weitschat
Sami Haddadin
Felix Huber
Alin Albu-Schäffer

Elastic robots have a distinct feature that makes them especially interesting to optimal control: their ability to mechanically store and release potential energy. However, solving any kind of optimal control problem for such highly nonlinear dynamics is feasible only numerically, i. e. offline. In turn, optimal solutions would only contribute a clear benefit for dynamic environments/tasks (apart from rather general insights), if they would be accessible/generalizable in real-time. In this paper, we propose a framework for executing near-optimal motions for elastic arms in real-time. We approach the problem as follows. First, we define a set of prototypical optimal control problems. These represent a reasonable set of motions that an intrinsically elastic robot arm is sought to execute. Exemplary, we solve the optimal control problem for some of these prototypes in a roughly covered task space. Then, we encode the resulting optimal trajectories in a dynamical system via Dynamic Movement Primitives (DMPs). Finally, a distance and cost function based metric forms the basis to generalize from the learned parameterizations to a new unsolved optimal control problem in real-time. In short, we intend to overcome the well known problems of optimal control and learning with associated generalization: being offline and being suboptimal, respectively.

Details

IROS Conference 2013 Conference Paper

First analysis and experiments in aerial manipulation using fully actuated redundant robot arm

Felix Huber
Konstantin Kondak
Kai Krieger
Dominik Sommer
Marc Schwarzbach
Maximilian Laiacker
Ingo Kossyk
Sven Parusel

In this paper we describe a system for aerial manipulation composed of a helicopter platform and a fully actuated seven Degree of Freedom (DoF) redundant industrial robotic arm. We present the first analysis of such kind of systems and show that the dynamic coupling between helicopter and arm can generate diverging oscillations with very slow frequency which we called phase circles. Based on the presented analysis, we propose a control approach for the whole system. The partial decoupling between helicopter and arm — which eliminates the phase circles — is achieved by means of special movement of robotic arm utilizing its redundant DoF. For the underlying arm control a specially designed impedance controller was proposed. In different flight experiments we showcase that the proposed kind of system type might be used in the future for practically relevant tasks. In an integrated experiment we demonstrate a basic manipulation task — impedance based grasping of an object from the environment underlaying a visual object tracking control loop.

Details

IROS Conference 2012 Conference Paper

Intrinsically elastic robots: The key to human like performance

Sami Haddadin
Felix Huber
Kai Krieger
Roman Weitschat
Alin Albu-Schäffer
Sebastian Wolf 0001
Werner Friedl
Markus Grebenstein

Intrinsically elastic robots, which technically implement some key characteristics of the human muskoskeletal system, have become a major research topic in nowadays robotics. These novel devices open up entirely new control approaches. They base on temporary storage of potential energy and its timed transformation into kinetic energy. In legged locomotion, such considerations have been a common tool for unveiling the respective fundamental physical processes. However, in arm control, elasticities were typically considered parasitic. In this video we outline our efforts in exploiting the inherent capabilities of intrinsically elastic robots in order to bring them closer to human performance. Instead of applying purely kinematic learing-by-demonstration approaches, which are certainly suboptimal, we argue for using model based techniques in order to optimally exploit the system dynamics such that highly dynamic motion and manipulation capabilities can be achieved. In particular, the explicit use of elasticities as temporary energy tanks can be fully exploited, if they are modeled adequately as an integral part of the mechanism. We also believe that such approaches can substantially contribute to the understanding of human motion biomechanics.

Details

ICRA Conference 2012 Conference Paper

Optimal control for exploiting the natural dynamics of Variable Stiffness robots

Sami Haddadin
Felix Huber
Alin Albu-Schäffer

In contrast to common rigid or actively compliant systems, Variable Stiffness Arms are capable of storing potential energy in their joint and convert it into kinetic energy, respectively speed. This capability is well known from humans and is a good example for the outstanding performance of biological systems. However, only since some years intrinsic compliance is considered as a key feature and not a drawback in robot design. Therefore, only very little work has been carried out on exploiting the natural dynamics of elastic arms for such explosive motion sequences. In this paper, we treat the problem of how to optimally achieve maximum link velocity at a given final time for Variable Stiffness Arms. We show that solutions to this problem lead to excitation motions, which enable the robot to move on the link side at much higher speed than on the motor side. In particular, the robot uses the dynamic transfer of elastic joint energy into link side kinetic energy for further acceleration. In our work we consider the practically relevant input and state constraints, and give experimental verification of the developed methods on the new DLR Hand-Arm system.

Details

IROS Conference 2012 Conference Paper

Optimal torque and stiffness control in compliantly actuated robots

David J. Braun
Florian Petit
Felix Huber
Sami Haddadin
Patrick van der Smagt
Alin Albu-Schäffer
Sethu Vijayakumar

Anthropomorphic robots that aim to approach human performance agility and efficiency are typically highly redundant not only in their kinematics but also in actuation. Variable-impedance actuators, used to drive many of these devices, are capable of modulating torque and passive impedance (stiffness and/or damping) simultaneously and independently. Here, we propose a framework for simultaneous optimisation of torque and impedance (stiffness) profiles in order to optimise task performance, tuned to the complex hardware and incorporating real-world constraints. Simulation and hardware experiments validate the viability of this approach to complex, state dependent constraints and demonstrate task performance benefits of optimal temporal impedance modulation.

Details