maplab: An Open Framework for Research in Visual-inertial Mapping and Localization aslteam 2026-09-22 | maplab: An Open Framework for Research in Visual-inertial Mapping and Localization
Continual Adaptation of Semantic Segmentation using Complementary 2D-3D Data Representations aslteam 2022-09-05 | by Jonas Frey, Hermann Blum, Francesco Milano, Roland Siegwart and Cesar Cadena.A previous version of this work was awarded the NeurIPS Robot Learning Workshop Best Paper Runner-up award.arxiv Link: arxiv.org/abs/2111.02156IEEE Link: ieeexplore.ieee.org/document/9874976Accepted for IEEE Robotics and Automation Letters (R-AL 2022)@inproceedings{frey2022continual,author={Jonas Frey and Hermann Blum and Francesco Milano and Roland Siegwart and Cesar Cadena},journal={IEEE Robotics and Automation Letters (R-AL 2022)},title={Continual Adaptation of Semantic Segmentation using Complementary 2D-3D Data Representations},year={2022}}Video by Jonas FreyAbstract:Semantic segmentation networks are usually pre-trained once and not updated during deployment. As a consequence, misclassifications commonly occur if the distribution of the training data deviates from the one encountered during the robot's operation. We propose to mitigate this problem by adapting the neural network to the robot's environment during deployment, without any need for external supervision. Leveraging complementary data representations, we generate a supervision signal, by probabilistically accumulating consecutive 2D semantic predictions in a volumetric 3D map. We then train the network on renderings of the accumulated semantic map, effectively resolving ambiguities and enforcing multi-view consistency through the 3D representation. In contrast to scene adaptation methods, we aim to retain the previously-learned knowledge, and therefore employ a continual learning experience replay strategy to adapt the network. Through extensive experimental evaluation, we show successful adaptation to real-world indoor scenes both on the ScanNet dataset and on in-house data recorded with an RGB-D sensor. Our method increases the segmentation accuracy on average by 9.9% compared to the fixed pre-trained neural network, while retaining knowledge from the pre-training dataset.
A Planning-and-Control Framework for Aerial Manipulation of Articulated Objects aslteam 2022-08-03 | This is the accompanying video of our RA-L paper titled "A Planning-and-Control Framework for Aerial Manipulation of Articulated Objects"ieeexplore.ieee.org/document/9830871Abstract:While the variety of applications for Aerial Manipulators (AMs) has increased over the last years, they are mostly limited to push-and-slide tasks. More complex manipulations of dynamic environments are poorly addressed and still require handcrafted designs of hardware, control, and trajectory planning. In this paper we focus on the active manipulation of articulated objects with AMs. We present a novel planning and control approach that allows the AM to execute complex interaction maneuvers with as little as possible priors given by the operator.Our framework combines sampling-based predictive control to generate pose trajectories with an impedance controller for compliant behaviours, applied to a fully-actuated flying platform. The framework leverages a physics engine to simulate the dynamics of the platform and the environment in order to find optimal motions to execute manipulation tasks.Experiments on two selected examples of pulling open a door and of turning a valve show the feasibility of the proposed approach.Reference:M. Brunner, G. Rizzi, M. Studiger, R. Siegwart and M. Tognon, "A Planning-and-Control Framework for Aerial Manipulation of Articulated Objects," in IEEE Robotics and Automation Letters, 2022, doi: 10.1109/LRA.2022.3191178
Adaptive Tank-based Control for Aerial Physical Interaction with Uncertain Dynamic Environments aslteam 2022-08-03 | This is the accompanying video of our RA-L paper titled "Adaptive Tank-based Control for Aerial Physical Interaction with Uncertain Dynamic Environments Using Energy-Task Estimation"ieeexplore.ieee.org/document/9826901Abstract:While aerial manipulation has witnessed noticeable growth as a field in the last decade, most works investigated forms of interaction with static and rigid environments only. Whenever dynamic environments were considered, the employed methods often relied on the knowledge of the model of the environment, which in most real applications cannot be obtained. In this work, we propose an adaptive controller for a fully actuated UAV performing stable and efficient physical interaction tasks with unmodeled and dynamic objects moving in unknown environments. We develop a passive time-varying impedance controller and wrench tracking controller, whose adaptable parameters allow us to minimize tracking error and instabilities during the execution of the interaction task. Robust stability is guaranteed by energy tanks, with the addition of a task-based formulation for adapting online the tank parameters in order to always provide the system with an adequate amount of energy. The control framework is validated both in simulations and experimentally by interacting with an unmodeled cart moving in passive time-varying environments, while subjected to unknown disturbances.Reference:F. Benzi, M. Brunner, M. Tognon, C. Secchi and R. Siegwart, "Adaptive Tank-based Control for Aerial Physical Interaction with Uncertain Dynamic Environments Using Energy-Task Estimation," in IEEE Robotics and Automation Letters, vol. 7, no. 4, pp. 9129-9136, Oct. 2022, doi: 10.1109/LRA.2022.3190074.PDF: https://www.research-collection.ethz.ch/handle/20.500.11850/561039
Under the Sand: A UAV with Ground Penetrating Synthetic Aperture Radar aslteam 2022-07-26 | This novel flying system combines an autonomous unmanned aerial vehicle with ground penetrating radar to detect buried objects such as landmines. The system stands out with sensor specific low altitude flight maneuvers and high accuracy position estimation. Experiments show radar detections of targets buried in sand.Please view our accompanying Field Robotics publication:doi.org/10.55417/fr.2022034All flying system software components are available open-source! Have a look at:github.com/ethz-asl/mav_findminegithub.com/ethz-asl/polygon_coverage_planninggithub.com/ethz-asl/ethz_piksi_rosgithub.com/ethz-asl/mav_gtsam_estimatorgithub.com/rikba/versavis/tree/feature/gnss_syncThis work was part of Project FindMine. A project initialized by the Urs Endress Foundation to develop new technologies for humanitarian demining. Have a look at the project page:ue-stiftung.org/findmineThe system is further developed within the FindMine gGmbH company! If you want to know more about humanitarian demining or support ongoing efforts please visit Fondation Suisse de Déminage (FSD)!https://www.fsd.ch/
Closed-Loop Next-Best-View Planning for Target-Driven Grasping aslteam 2022-07-22 | Accompanying video of our IROS 2022 Submission.Paper: http://arxiv.org/abs/2207.10543Code: github.com/ethz-asl/active_graspAbstract: Picking a specific object from clutter is an essential component of many manipulation tasks. Partial observations often require the robot to collect additional views of the scene before attempting a grasp. This paper proposes a closed-loop next-best-view planner that drives exploration based on occluded object parts. By continuously predicting grasps from an up-to-date scene reconstruction, our policy can decide online to finalize a grasp execution or to adapt the robot's trajectory for further exploration. We show that our reactive approach decreases execution times without loss of grasp success rates compared to common camera placements and handles situations where the fixed baselines fail.
Fast and Compute-efficient Sampling-based Local Exploration Planning via Distribution Learning aslteam 2022-07-04 | Paper: arxiv.org/abs/2202.13715Code: github.com/ethz-asl/cvae_exploration_planningAbstract:Exploration is a fundamental problem in robotics. While sampling-based planners have shown high performance and robustness, they are oftentimes compute intensive and can exhibit high variance. To this end, we propose to learn both components of sampling-based exploration. We present a method to directly learn an underlying informed distribution of views based on the spatial context in the robot's map, and further explore a variety of methods to also learn the information gain of each sample. We show in thorough experimental evaluation that our proposed system improves exploration performance by up to 28\% over classical methods, and find that learning the gains in addition to the sampling distribution can provide favorable performance vs. compute trade-offs for compute-constrained systems.We demonstrate in simulation and on a low-cost mobile robot that our system generalizes well to varying environments.
Avalmapper: Monitoring avalanches with long endurance UAVs aslteam 2022-06-08 | The Avalmapper project aims to improve avalanche monitoring by utilizing long endurance UAVs. Using these UAVs high-resolution avalanche occurrence data over affected regions in the Swiss Alps can be obtained. This video shows footage from winter test flights of 2021Contact: Jaeyoung Lim jalim@ethz.chThe avalmapper project is a collaboration between the Autonomous Systems Lab (ASL) at ETHZurich and SLF.For more information of the project visit: https://avalmapper.ethz.ch/
Aerial Layouting: Design of a Compliant and Actuated End-Effector for Precise Marking on Ceilings aslteam 2022-06-07 | We developed a self-contained end-effector for layouting on construction sites with aerial robots! The end-effector achieves high accuracy through the use of multiple contact points, compliance, and actuation. Find out more about the system design in our paper: http://www.roboticsproceedings.org/rss18/p073.pdf Contact: Christian Lanegger: christian.lanegger@mavt.ethz.chAbstract: Aerial robots have demonstrated impressive feats of precise control, such as dynamic flight through openings or highly complex choreographies. Despite the accuracy needed for these tasks, there are problems that require levels of precision that are challenging to achieve today. One such problem is aerial interaction. Advances in aerial robot design and control have made such contact-based tasks possible and opened up research into challenging real-world tasks, including contact-based inspection. However, while centimetre accuracy is sufficient and achievable for inspection tasks, the positioning accuracy needed for other problems, such as layouting on construction sites or general push-and-slide tasks, is millimetres. To achieve such high precision, we propose a new aerial system composed of an aerial vehicle equipped with a novel ``smart'' end-effector leveraging a stability-optimized Gough-Stewart mechanism. We present its design process and features incorporating the principles of compliance, multiple contact points, actuation, and self-containment.In experiments, we verify that the design choices made for our novel end-effector are necessary to obtain the desired positioning precision. Furthermore, we demonstrate that we can reliably mark lines on ceilings with millimetre accuracy without the need for precise modeling or sophisticated control of the aerial robot.
Power-based Safety Layer for Aerial Vehicles in Physical Interaction using Lyapunov Exponents aslteam 2022-05-25 | This is the accompanying video of our RA-L paper titled "Power-based Safety Layer for Aerial Vehicles in Physical Interaction using Lyapunov Exponents"ieeexplore.ieee.org/document/9780016Abstract:As the performance of autonomous systems increases, safety concerns arise, especially when operating in non-structured environments. To deal with these concerns, this work presents a safety layer for mechanical systems that detects and responds to unstable dynamics caused by external disturbances. The safety layer is implemented independently and on top of already present nominal controllers, like pose or wrench tracking, and limits power flow when the system's response would lead to instability. This approach is based on the computation of the Largest Lyapunov Exponent (LLE) of the system's error dynamics, which represent a measure of the dynamics' divergence or convergence rate. By actively computing this metric, divergent and possibly dangerous system behaviors can be promptly detected. The LLE is then used in combination with Control Barrier Functions (CBFs) to impose power limit constraints on a jerk controlled system. The proposed architecture is experimentally validated on an Omnidirectional Micro Aerial Vehicle (OMAV) both in free flight and interaction tasks.Reference:E. Cuniato, N. R. J. Lawrance, M. Tognon and R. Siegwart, "Power-based Safety Layer for Aerial Vehicles in Physical Interaction using Lyapunov Exponents," in IEEE Robotics and Automation Letters, doi: 10.1109/LRA.2022.3176959.PDF: https://www.research-collection.ethz.ch/handle/20.500.11850/548959Affiliations:All authors are with the Autonomous Systems Lab, ETH Zurich, 8092 Switzerland.
[ICRA22] Collaborative Robot Mapping using Spectral Graph Analysis aslteam 2022-05-12 | ArXiv: arxiv.org/abs/2203.00308Github: github.com/ethz-asl/fgspContact: Lukas Bernreiter berlukas@ethz.chAbstract: In this paper, we deal with the problem of creating globally consistent pose graphs in a centralized multi-robot SLAM framework. For each robot to act autonomously, individual onboard pose estimates and maps are maintained, which are then communicated to a central server to build an optimized global map.However, inconsistencies between onboard and server estimates can occur due to onboard odometry drift or failure. Furthermore, robots do not benefit from the collaborative map if the server provides no feedback in a computationally tractable and bandwidth-efficient manner.Motivated by this challenge, this paper proposes a novel collaborative mapping framework to enable accurate global mapping among robots and server. In particular, structural differences between robot and server graphs are exploited at different spatial scales using graph spectral analysis to generate necessary constraints for the individual robot pose graphs.The proposed approach is thoroughly analyzed and validated using several real-world multi-robot field deployments where we show improvements in the onboard system up to 90%.
Embodied Active Domain Adaptation for Semantic Segmentation via Informative Path Planning aslteam 2022-03-21 | Paper: doi.org/10.48550/arXiv.2203.00549Code: github.com/ethz-asl/active_learning_for_segmentationAbstract:This work presents an embodied agent that can adapt its semantic segmentation network to new indoor environments in a fully autonomous way. Because semantic segmentation networks fail to generalize well to unseen environments, the agent collects images of the new environment which are then used for self-supervised domain adaptation. We formulate this as an informative path planning problem, and present a novel information gain that leverages uncertainty extracted from the semantic model to safely collect relevant data. As domain adaptation progresses, these uncertainties change over time and the rapid learning feedback of our system drives the agent to collect different data. Experiments show that our method adapts to new environments faster and with higher final performance compared to an exploration objective, and can successfully be deployed to real-world environments on physical robots.
Energy Tank-Based Policies for Robust Aerial Physical Interaction with Moving Objects aslteam 2022-03-09 | This is the accompanying video for our ICRA publication titled "Energy Tank -Based Policies for Robust Aerial Physical Interaction with Moving Objects".Abstract:Although manipulation capabilities of aerial robots greatly improved in the last decade, only few works addressed the problem of aerial physical interaction with dynamic environments, proposing strongly model-based approaches. However, in real scenarios, modeling the environment with high accuracy is often impossible.In this work, we aim at developing a control framework for OMAVs for reliable physical interaction tasks with articulated and movable objects in the presence of possibly unforeseen disturbances, and without relying on an accurate model of the environment. Inspired by previous applications of energy-based controllers for physical interaction, we propose a passivity-based impedance and wrench tracking controller in combination with a momentum-based wrench estimator.This is combined with an energy-tank framework to guarantee the stability of the system, while energy and power flow-based adaptation policies are deployed to enable safe interaction with any type of passive environment.The control framework provides formal guarantees of stability, which is validated in practice considering the challenging task of pushing a cart of unknown mass, moving on a surface of unknown friction, as well as subjected to unknown disturbances.For this scenario, we present, evaluate and discuss three different policies.Affiliations:All authors are with the Autonomous Systems Lab (ASL), Department of Mechanical and Process Engineering, ETH Zurich, 8092 Zurich, Switzerland.
Towards 6DoF Bilateral Teleoperation of an Omnidirectional Aerial Vehicle for Aerial Interaction aslteam 2022-03-07 | This is the accompanying video of our ICRA paper titled "Towards 6DoF Bilateral Teleoperation of an Omnidirectional Aerial Vehicle for Aerial Physical Interaction"PDF: https://www.research-collection.ethz.ch/handle/20.500.11850/535634Abstract:Bilateral teleoperation offers an intriguing solution towards shared autonomy with aerial vehicles in contact-based inspection and manipulation tasks. Omnidirectional aerial robots allow for full pose operations, making them particularly attractive in such tasks. Naturally, the question arises whether standard bilateral teleoperation methodologies are suitable for use with these vehicles. In this work, a fully decoupled 6DoF bilateral teleoperation framework for aerial physical interaction is designed and tested for the first time. The method is based on the well established rate control, recentering and interaction force feedback policy. However, practical experiments evince the difficulty of performing decoupled motions in a single axis only. As such, this work shows that the trivial extension of standard methods is insufficient for omnidirectional teleoperation, due to the operator’s physical inability to properly decouple all input DoFs. This suggests that further studies on enhanced haptic feedback are necessary.Reference:Mike Allenspach, Nicholas Lawrance, Marco Tognon, Roland Siegwart "Towards 6DoF Bilateral Teleoperation of an Omnidirectional Aerial Vehicle for Aerial Physical Interaction" IEEE International Conference on Robotics and Automation (ICRA), 2022Affiliations:All authors are with the Autonomous Systems Lab, ETH Zurich, 8092 Switzerland.
Panoptic Multi-TSDFs: Online Multi-resolution Mapping with Long-term Dynamic Scene Consistency aslteam 2022-02-22 | Panoptic Mapping is our latest framework for flexible, object-based multi-resolution volumetric mapping in changing scenes! Find out more in our paper and on github!ArXiv: arxiv.org/abs/2109.10165Github: github.com/ethz-asl/panoptic_mappingContact: Lukas Schmid schmluk@ethz.chAbstract: For robotic interaction in environments shared with other agents, access to volumetric and semantic maps of the scene is crucial. However, such environments are inevitably subject to long-term changes, which the map needs to account for. We thus propose panoptic multi-TSDFs as a novel representation for multi-resolution volumetric mapping in changing environments. By leveraging high-level information for 3D reconstruction, our proposed system allocates high resolution only where needed. Through reasoning on the object level, semantic consistency over time is achieved. This enables our method to maintain up-to-date reconstructions with high accuracy while improving coverage by incorporating previous data. We show in thorough experimental evaluation that our map can be efficiently constructed, maintained, and queried during online operation, and that the presented approach can operate robustly on real depth sensors using non-optimized panoptic segmentation as input.
Human-State-Aware Controller for a Tethered Aerial Robot Guiding a Human by Physical Interaction aslteam 2022-02-01 | This is the accompanying video of our RA-L paper titled "Human-State-Aware Controller for a Tethered Aerial Robot Guiding a Human by Physical Interaction"ieeexplore.ieee.org/document/9684670Abstract:With the rapid development of Aerial Physical Interaction, the possibility to have aerial robots physically interacting with humans is attracting a growing interest. In one of our previous works, we considered one of the first systems in which a human is physically connected to an aerial vehicle by a cable. There, we developed a compliant controller that allows the robot to pull the human toward a desired position using forces only as an indirect communication-channel. However, this controller is based on the robot-state only, which makes the system not adaptable to the human behavior, and in particular to their walking speed. This reduces the effectiveness and comfort of the guidance when the human is still far from the desired point. In this paper, we formally analyze the problem and propose a human-state-aware controller that includes a human’s velocity feedback. We theoretically prove and experimentally show that this method provides a more consistent guiding force which enhances the guiding experience.Reference:Mike Allenspach, Yash Vyas, Matthias Rubio, Roland Siegwart, and Marco Tognon "Human-State-Aware Controller for a Tethered Aerial Robot Guiding a Human by Physical Interaction" IEEE Robotics and Automation Letters, 2021PDF: https://www.research-collection.ethz.ch/bitstream/handle/20.500.11850/530266/final_version.pdf?sequence=1&isAllowed=yAffiliations:All authors are with the Autonomous Systems Lab, ETH Zurich, 8092 Switzerland.
This is Ascento Pro aslteam 2022-01-19 | Bigger, faster, stronger: This is Ascento Pro! Ascento Robotics newest creation can climb full flights of stairs, drive at up to 12km/h and all this for up to 8h per battery charge. Furthermore, it now features full autonomy.For more information visit https://www.ascento.ch or check out the YouTube channel of @Ascento Robotics
DARPA SubT Challenge - Technology Overview by Our Winning Team CERBERUS aslteam 2021-12-08 | The DARPA SubT Challenge: subtchallenge.comTeam CERBERUS is an international collaboration between several labs, for a full list see: http://www.subt-cerberus.org/cerberus-team.htmlFor videos of the runs, see DARPAtv: youtube.com/playlist?list=PL6wMum5UsYvYpbhQALOcbhzXYTt3qnzqAA competition summary by WaPo: washingtonpost.com/magazine/2021/11/10/darpa-robot-competition
Continual Learning of Semantic Segmentation using Complementary 2D-3D Data Representations aslteam 2021-11-04 | Paper: arxiv.org/abs/2111.02156Semantic segmentation networks are usually pre-trained and not updated during deployment. As a consequence, misclassifications commonly occur if the distribution of the training data deviates from the one encountered during the robot's operation. We propose to mitigate this problem by adapting the neural network to the robot's environment during deployment, without any need for external supervision. Leveraging complementary data representations, we generate a supervision signal, by probabilistically accumulating consecutive 2D semantic predictions in a volumetric 3D map. We then retrain the network on renderings of the accumulated semantic map, effectively resolving ambiguities and enforcing multi-view consistency through the 3D representation. To preserve the previously-learned knowledge while performing network adaptation, we employ a continual learning strategy based on experience replay. Through extensive experimental evaluation, we show successful adaptation to real-world indoor scenes both on the ScanNet dataset and on in-house data recorded with an RGB-D sensor. Our method increases the segmentation performance on average by 11.8% compared to the fixed pre-trained neural network, while effectively retaining knowledge from the pre-training dataset.
Dynamic End Effector Tracking with an Omnidirectional Parallel Aerial Manipulator aslteam 2021-08-16 | This is the accompanying video of our RAL paper titled "Dynamic End Effector Tracking with an Omnidirectional Parallel Aerial Manipulator"ieeexplore.ieee.org/document/9507086Abstract:To address the challenge of precise, dynamic and versatile aerial manipulation, we present an aerial manipulation platform consisting of a parallel 3-DOF manipulator mounted to an omnidirectional tilt-rotor aerial vehicle. The general modeling of a parallel manipulator on an omnidirectional floating base is presented, which motivates the optimization and detailed design of the aerial manipulator parameters and components. Inverse kinematic control of the manipulator is coupled to the omnidirectional base pose controller with a dynamic compensation term, going beyond common decoupled approaches. This presents a baseline for the control of redundant omnidirectional aerial manipulators. Experimental flights show the advantages of an active manipulator vs. a fixed arm for disturbance rejection and end effector tracking performance, as well as the practical limitations of the dynamic compensation term for fast endeffector trajectories. The results motivate future studies for precise and dynamic aerial manipulation.Reference:Karen Bodie, Marco Tognon, and Roland Siegwart "Dynamic End Effector Tracking with an Omnidirectional Parallel Aerial Manipulator" IEEE Robotics and Automation Letters, 2021PDF: https://www.research-collection.ethz.ch/bitstream/handle/20.500.11850/499753/main.pdf?sequence=1&isAllowed=yAffiliations: K. Bodie, M. Tognon, and R. Siegwart are with the Autonomous SystemsLab, ETH Zurich, 8092 Switzerland
Precise Robot Localization in Architectural 3D Plans aslteam 2021-07-30 | Arxiv Paper: arxiv.org/abs/2006.05137This paper presents a localization system for mobile robots enabling precise localization in inaccurate building models. The approach leverages local referencing to counteract inherent deviations between as-planned and as-built data for locally accurate registration. We further fuse a novel camera-based robust outlier detector with LiDAR data to reject a wide range of outlier measurements from clutter, dynamic objects, and sensor failures. We evaluate the proposed approach on a mobile robot in a challenging real world site. In presence of clutter and model deviations, our system reduces the localization error by at least 30%.
Dipper: A Dynamically Transitioning Aerial-Aquatic Unmanned Vehicle aslteam 2021-07-05 | Dipper is an aerial-aquatic hybrid vehicle, capable of controlled motion in air and underwater. Dipper is a lightweight fixed-wind UAV with actively swept wings that achieves dynamic transitions between the two media. The vehicle has only one main propulsion motor, and uses a novel clutch system to engage either the front tractor propeller for flight in air, or the rear ship's screw propeller for underwater propulsion. More details can be found on our project page at https://dipper.ethz.ch/.This is video is associated with the following publication:F Rockenbauer, S Jeger, L Beltran, M Berger, M Harms, N Kaufmann, M Rauch, M Reinders, N Lawrance, T Stastny, R Siegwart. Dipper: A Dynamically Transitioning Aerial-Aquatic Unmanned Vehicle. Robotics: Science and Systems, 2021.Link to paper: http://www.roboticsproceedings.org/rss17/p048.pdf
Transitioning between Outdoors and GPS-Denied Environments with Aerial Vehicles aslteam 2021-06-22 | Indoor/Outdoor transition and flight close to structure still poses a challenge to aerial robots using GPS for navigation. In this work we address the problem by integrating high precision GPS measurements into our visual inertial odometry pipeline. Using the uncertainty of the GPS measurements we are able to consider only accurate GPS updates and reject unreliable ones. This allows us to obtain a smooth and accurate state estimate throughout a whole flight in a environment where GPS navigation only would not be possible.Relevant Paper: onlinelibrary.wiley.com/doi/abs/10.1002/rob.21824Code: VIO: github.com/ethz-asl/rovio GPS: github.com/ethz-asl/ethz_piksi_rosPlatform: voliro.com
GLocal: Efficient, Complete, and Safe Volumetric Exploration under Severe Odometry Drift aslteam 2021-06-18 | GLocal is our latest framework for fully autonomous volumetric exploration subject to state estimation drift! Find out more in our paper and on github!Paper: IEEE: ieeexplore.ieee.org/abstract/document/9387110 ArXiv: arxiv.org/abs/2010.09859Code: GLocal: github.com/ethz-asl/glocal_exploration Simulator: github.com/ethz-asl/unreal_airsimContact: Lukas Schmid: schmluk@mavt.ethz.ch Victor Reijgwart: victor.reijgwart@mavt.ethz.chAbstract: Exploration is a fundamental problem in robot autonomy. A major limitation, however, is that during exploration robots oftentimes have to rely on on-board systems alone for state estimation, accumulating significant drift over time in large environments. Drift can be detrimental to robot safety and exploration performance. In this work, a submap-based, multi-layer approach for both mapping and planning is proposed to enable safe and efficient volumetric exploration of large scale environments despite odometry drift. The central idea of our approach combines local (temporally and spatially) and global mapping to guarantee safety and efficiency. Similarly, our planning approach leverages the presented map to compute global volumetric frontiers in a changing global map and utilizes the nature of exploration dealing with partial information for efficient local and global planning. The presented system is thoroughly evaluated and shown to outperform state of the art methods even under drift-free conditions. Our system, termed GLocal , is made available open source.Background Music:Time and Space by Keys of Moon | soundcloud.com/keysofmoonAttribution 4.0 International (CC BY 4.0)creativecommons.org/licenses/by/4.0Music promoted by chosic.com
Pixel wise Anomaly Detection in Complex Driving Scenes aslteam 2021-06-01 | CVPR Talk for our paper "Pixel-Wise Anomaly Detection in Complex Driving Scenes"Paper Link: arxiv.org/abs/2103.05445Code Link: github.com/giandbt/synboostAbstract:The inability of state-of-the-art semantic segmentation methods to detect anomaly instances hinders them from being deployed in safety-critical and complex applications, such as autonomous driving. Recent approaches have focused on either leveraging segmentation uncertainty to identify anomalous areas or re-synthesizing the image from the semantic label map to find dissimilarities with the input image. In this work, we demonstrate that these two methodologies contain complementary information and can be combined to produce robust predictions for anomaly segmentation. We present a pixel-wise anomaly detection framework that uses uncertainty maps to improve over existing re-synthesis methods in finding dissimilarities between the input and generated images. Our approach works as a general framework around already trained segmentation networks, which ensures anomaly detection without compromising segmentation accuracy, while significantly outperforming all similar methods. Top-2 performance across a range of different anomaly datasets shows the robustness of our approach to handling different anomaly instances.
Aerial Manipulation with an Omnidirectional Flying Platform aslteam 2021-05-25 | This video showcases our current research to address the challenges of aerial manipulation with omnidirectional flying robots at ETH Zurich's Autonomous systems lab. This topic connects several different topics of active research at our institute, including design and control of omnidirectional aerial manipulators, planning frameworks, and grasp detection.For further details on the content of this video, links to the related research papers are listed below.----- Paper References -----[1] Allenspach, M., Bodie, K., Brunner, M., Rinsoz, L., Taylor, Z., Kamel, M., Siegwart, R. and Nieto, J., 2020. "Design and optimal control of a tiltrotor micro-aerial vehicle for efficient omnidirectional flight.", The International Journal of Robotics Research, 39(10-11), pp.1305-1325.link: arxiv.org/pdf/2003.09512.pdf[2] Zhang, W., Brunner, M., Ott, L., Kamel, M., Siegwart, R. and Nieto, J., 2020. "Learning dynamics for improving control of overactuated flying systems.", IEEE Robotics and Automation Letters, 5(4), pp.5283-5290.link: arxiv.org/pdf/2006.13153.pdf[3] Zhang, W., Ott, L., Tognon, M., Siegwart, R. and Nieto, J., 2021. "Active Model Learning using Informative Trajectories for Improved Closed-Loop Control on Real Robots." arXiv preprint arXiv:2101.08100.link: arxiv.org/pdf/2101.08100.pdf[4] Bodie, K., Brunner, M., Pantic, M., Walser, S., Pfändler, P., Angst, U., Siegwart, R. and Nieto, J., 2019. "An omnidirectional aerial manipulation platform for contact-based inspection.", Robotics: Science and System XV, 15.link: arxiv.org/pdf/1905.03502.pdf [5] Bodie, K., Tognon, M. and Siegwart, R. "Dynamic End Effector Tracking with an Omnidirectional Parallel Aerial Manipulator." (under review, link to follow)[6] Breyer, M., Chung, J.J., Ott, L., Siegwart, R. and Nieto, J., 2021. "Volumetric Grasping Network: Real-time 6 DOF Grasp Detection in Clutter.", arXiv preprint arXiv:2101.01132.link: arxiv.org/pdf/2101.01132.pdf[7] Pantic, M., Ott, L., Cadena, C., Siegwart, R. and Nieto, J., 2021. "Mesh Manifold based Riemannian Motion Planning for Omnidirectional Micro Aerial Vehicles.", IEEE Robotics and Automation Letters, 6(3), pp.4790-4797.link: arxiv.org/pdf/2102.10313.pdf[8] Allenspach, M., Lawrance, N. , Tognon, M. and Siegwart, R. "6 DOF Teleoperation of an Omnidirectional Aerial Vehicle." (under review, link to follow)
Dynamic Object Aware LiDAR SLAM based on Automatic Generation of Training Data aslteam 2021-05-20 | Highly dynamic environments, with moving objects such as cars or humans, can pose a performance challenge for LiDAR SLAM systems that assume largely static scenes. To overcome this challenge and support the deployment of robots in real world scenarios, we propose a complete solution for a dynamic object aware LiDAR SLAM algorithm. This is achieved by leveraging a real-time capable neural network that can detect dynamic objects, thus allowing our system to deal with them explicitly. To efficiently generate the necessary training data which is key to our approach, we present a novel end-to-end occupancy grid based pipeline that can automatically label a wide variety of arbitrary dynamic objects. Our solution can thus generalize to different environments without the need for expensive manual labeling and at the same time avoids assumptions about the presence of a predefined set of known objects in the scene. Using this technique, we automatically label over 12000 LiDAR scans collected in an urban environment with a large amount of pedestrians and use this data to train a neural network, achieving an average segmentation IoU of 0.82. We show that explicitly dealing with dynamic objects can improve the LiDAR SLAM odometry performance by 39.6% while yielding maps which better represent the environments. Paper preprint: arxiv.org/abs/2104.03657Dataset: https://projects.asl.ethz.ch/datasets/doals
IAN: Multi-Behavior Navigation Planning for Robots in Real, Crowded Environments (IROS 2020 Full) aslteam 2021-05-18 | This is the full presentation video as shown during the IROS 2020 conference.paper: http://ras.papercept.net/images/temp/IROS/files/1703.pdfcode: github.com/ethz-asl/interaction_actions_for_navigation
TSDF++: A Multi-Object Formulation for Dynamic Object Tracking and Reconstruction aslteam 2021-05-18 | By Margarita Grinvald, Federico Tombari, Roland Siegwart, and Juan Nieto.Abstract - The ability to simultaneously track and reconstruct multiple objects moving in the scene is of the utmost importance for robotic tasks such as autonomous navigation and interaction. Virtually all of the previous attempts to map multiple dynamic objects have evolved to store individual objects in separate reconstruction volumes and track the relative pose between them. While simple and intuitive, such formulation does not scale well with respect to the number of objects in the scene and introduces the need for an explicit occlusion handling strategy. In contrast, we propose a map representation that allows maintaining a single volume for the entire scene and all the objects therein. To this end, we introduce a novel multi-object TSDF formulation that can encode multiple object surfaces at any given location in the map. In a multiple dynamic object tracking and reconstruction scenario, our representation allows maintaining accurate reconstruction of surfaces even while they become temporarily occluded by other objects moving in their proximity. We evaluate the proposed TSDF++ formulation on a public synthetic dataset and demonstrate its ability to preserve reconstructions of occluded surfaces when compared to the standard TSDF map representation.The paper has been accepted to the 2021 IEEE International Conference on Robotics and Automation (ICRA) in Xi'an, China.Paper: arxiv.org/abs/2105.07468Code: github.com/ethz-asl/tsdf-plusplus
Active Model Learning using Informative Trajectories for Improved Closed-Loop Control on Real Robots aslteam 2021-05-14 | This is the accompanying video for our ICRA conference paper titled "Active Model Learning using Informative Trajectories for Improved Closed-Loop Control on Real Robots" arxiv.org/abs/2101.08100Abstract: Model-based controllers on real robots require accurate knowledge of the system dynamics to perform optimally. For complex dynamics, first-principles modeling is not sufficiently precise, and data-driven approaches can be leveraged to learn a statistical model from real experiments. However, the efficient and effective data collection for such a data-driven system on real robots is still an open challenge. This paper introduces an optimization problem formulation to find an informative trajectory that allows for efficient data collection and model learning. We present a sampling-based method that computes an approximation of the trajectory that minimizes the prediction uncertainty of the dynamics model. This trajectory is then executed, collecting the data to update the learned model. In experiments we demonstrate the capabilities of our proposed framework when applied to a complex omnidirectional flying vehicle with tiltable rotors. Using our informative trajectories results in models which outperform models obtained from non-informative trajectory by 13.3\% with the same amount of training data. Furthermore, we show that the model learned from informative trajectories generalizes better than the one learned from non-informative trajectories, achieving better tracking performance on different tasks.Affiliations:All authors are with the Autonomous Systems Lab, Department of Mechanical and Process Engineering, ETH Zurich, 8092 Zurich, Switzerland.
Self-Improving Semantic Perception on a Construction Robot aslteam 2021-05-07 | Paper: arxiv.org/abs/2105.01595We propose a novel robotic system that can improve its semantic perception during deployment. Our system tightly couples multi-sensor perception and localisation to continuously learn from self-supervised pseudo labels.
Online Synchronization of Building Model for On-Site Mobile Robotic Construction aslteam 2021-05-04 | Paper: https://www.research-collection.ethz.ch/bitstream/handle/20.500.11850/455152/1/ISARC2020_Paper_213.pdfThis research presents a novel method for a data flow that synchronizes building information with the robot map and updates building components to their "as-built" states, in order to facilitate an on-site mobile construction process. Our experiments showcase mobile mapping and localization of a robotic platform featuring segmentation, plane association and quantitative evaluation of deviations. For the users of the on-site mobile robotic system, we present a suitable interface that allows for task level commanding and the selection of target and reference building components (i.e. walls, floor, ceiling). Additionally, this interface seamlessly integrates the online workflow between building construction and the robot map, updating the target building components to their "as-built" states in real time and providing a visual representation of additional task-specific attributes for building components in the robot map, in addition to geometries. This is presented as a first step toward integrating users of the system into the proposed robotic workflow to develop decision-making strategies for fitting building tasks to local references on-site.
Direct Force and Pose NMPC with Multiple Interaction Modes for Aerial Push-and-Slide Operations aslteam 2021-03-30 | We present a nonlinear model predictive controller (NMPC) for a fully actuated aerial manipulator to track a hybrid force and pose trajectory at the end-effector in an aerial interaction task. A force sensor at the end-effector is used to detect contact and to directly control the interaction force. We propose an approach for automatic transition between three operation modes which reflect the state of contact constraints, including free flight and two modes for force control based on static or dynamic friction at the end-effector. This division into three modes allows for different mode-specific controller tunings to optimize the desired performance throughout an interaction task.
Mesh Manifold based Riemannian Motion Planning for Omnidirectional Micro Aerial Vehicles aslteam 2021-02-23 | Preprint: arxiv.org/abs/2102.10313This paper presents a novel on-line path planning method that enables aerial robots to interact with surfaces. We present a solution to the problem of finding trajectories that drive a robot towards a surface and move along it. Triangular meshes are used as a surface map representation that is free of fixed discretization and allows for very large workspaces.We propose to leverage planar parametrization methods to obtain a lower-dimensional topologically equivalent representation of the original surface.Furthermore, we interpret the original surface and its lower-dimensional representation as manifold approximations that allow the use of Riemannian Motion Policies (RMPs), resulting in an efficient, versatile, and elegant motion generation framework. We compare against several Rapidly-exploring Random Tree (RRT) planners, a customized CHOMP variant, and the discrete geodesic algorithm. Using extensive simulations on real-world data we show that the proposed planner can reliably plan high-quality near-optimal trajectories at minimal computational cost. The accompanying multimedia attachment demonstrates feasibility on a real OMAV. The obtained paths show less than 10% deviation from the theoretical optimum while facilitating reactive re-planning at kHz refresh rates, enabling flying robots to perform motion planning for interaction with complex surfaces.Michael Pantic, Lionel Ott, Cesar Cadena, Roland Siegwart, and Juan Nieto
Hough2Map - Iterative Event-based Hough Transform for High-Speed Railway Mapping aslteam 2021-02-16 | To cope with the growing demand for transportation on the railway system, accurate, robust, and high-frequency positioning is required to enable a safe and efficient utilization of the existing railway infrastructure. As a basis for a localization system we propose a complete on-board mapping pipeline able to map robust meaningful landmarks, such as poles from power lines, in the vicinity of the vehicle. Such poles are good candidates for reliable and long-term landmarks even through difficult weather conditions or seasonal changes. To address the challenges of motion blur and illumination changes in railway scenarios we employ a Dynamic Vision Sensor, a novel event-based camera. Using a sideways oriented on-board camera, poles appear as vertical lines. To map such lines in a real-time event stream, we introduce Hough2Map, a novel consecutive iterative event-based Hough transform framework capable of detecting, tracking, and triangulating close-by structures. We demonstrate the mapping reliability and accuracy of Hough2Map on real-world data in typical usage scenarios and evaluate using surveyed infrastructure ground truth maps. Hough2Map achieves a detection reliability of up to 92% and a mapping root mean square error accuracy of 1.1518m.Paper preprint: arxiv.org/abs/2102.08145Code: github.com/ethz-asl/Hough2Map
VGN: Real-Time 6 DOF Grasp Detection in Clutter aslteam 2021-01-05 | Supplementary video of our CoRL 2020 submission, "Volumetric Grasping Network: Real-Time 6 DOF Grasp Detection in Clutter".arXiv: http://arxiv.org/abs/2101.01132Github: github.com/ethz-asl/vgnAbstract - General robot grasping in clutter requires the ability to synthesize grasps that work for previously unseen objects and that are also robust to physical interactions, such as collisions with other objects in the scene. In this work, we design and train a network that predicts 6 DOF grasps from 3D scene information gathered from an on-board sensor such as a wrist-mounted depth camera. Our proposed Volumetric Grasping Network (VGN) accepts a Truncated Signed Distance Function (TSDF) representation of the scene and directly outputs the predicted grasp quality and the associated gripper orientation and opening width for each voxel in the queried 3D volume. We show that our approach can plan grasps in only 10 ms and is able to clear 92% of the objects in real-world clutter removal experiments without the need for explicit collision checking. The real-time capability opens up the possibility for closed-loop grasp planning, allowing robots to handle disturbances, recover from errors and provide increased robustness.
Autonomous Christmas Lab 2020 - Short Version aslteam 2020-12-22 | The Autonomous Systems Lab at ETH Zurich wishes everyone a Merry Christmas and a Happy 2021Music:Canon and Variation - Twin MusicoomHolly Dazed - RKVCPrizefighter - Norma Rockwell
Autonomous Christmas Lab 2020 aslteam 2020-12-20 | The Autonomous Systems Lab at ETH Zurich wishes everyone a Merry Christmas and a Happy 2021Music:Canon and Variation - Twin MusicoomHolly Dazed - RKVCPrizefighter - Norma Rockwell
Fault-Tolerant Flight Control of a VTOL Tailsitter UAV aslteam 2020-12-02 | By Silvan Fuhrer, Sebastian Verling, Thomas Stastny, and Roland SiegwartAbstract -- Compared to other vertical take-off and landing (VTOL) systems, a tailsitter minimizes the number of actuators and moving parts necessary. The downside of having a minimalistic actuation is its inherent low fault-tolerance. The failure of an actuator usually results in a loss of controllability, resulting in a crash. In this paper we analyze possible actuator failures and the constraints they pose on the capabilities of the system. We further present light-weight adaptations to the nominal flight controller to make it fault-tolerant. The fault-tolerant controller is implemented on a small tailsitter VTOL aircraft and adjusted to the system by means of extensive experimental studies. Finally, the capabilities and performance under failures are demonstrated and analyzed.This work was published in the 2019 International Conference on Robotics and Automation (ICRA).Paper: ieeexplore.ieee.org/document/8793467
IAN: Multi-Behavior Navigation Planning for Robots in Real, Crowded Environments aslteam 2020-11-16 | For more information, our paper is available at http://ras.papercept.net/images/temp/IROS/files/1703.pdfIAN is on github! github.com/ethz-asl/interaction_actions_for_navigationThe track is After You've Gone, performed by the U.S. Coast Guard Band (public domain)
Attitude- and Cruise Control of a VTOL Tiltwing UAV aslteam 2020-11-15 | By David Rohr, Thomas Stastny, Sebastian Verling and Roland SiegwartAbstract—This paper presents the mathematical modeling, controller design, and flight-testing of an over-actuated Vertical Take-off and Landing (VTOL) tiltwing Unmanned Aerial Vehicle (UAV). Based on simplified aerodynamics and first-principles, a dynamical model of the UAV is developed which captures key aerodynamic effects including propeller slipstream on the wing and post-stall characteristics of the airfoils. The model-based steady-state flight envelope and the corresponding trim-actuation is analyzed and the overactuation of the UAV solved by optimizing for, e.g., power-optimal trims. The developed control system is composed of two controllers: First, a low-level attitude controller based on dynamic inversion and a daisy-chaining approach to handle allocation of redundant actuators. Secondly, a higher-level cruise controller to track a desired vertical velocity. It is based on a linearization of the system and look-up tables to determine the strong and nonlinear variation of the trims throughout the flight-envelope. We demonstrate the performance of the control- system for all flight phases (hover, transition, cruise) in extensive flight-tests.The paper has been accepted to IEEE Robotics and Automation Letters (RA-L) and IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2019 in Macau.Paper: ieeexplore.ieee.org/document/8703808arXiv: arxiv.org/abs/1903.10623This work was supported by Dufour Aerospace (https://dufour.aero)
Freetures: Localization in Signed Distance Function Maps aslteam 2020-10-19 | Localization of a robotic system within a previously mapped environment is important for reducing estimation drift and for reusing previously built maps. Existing techniques for geometry-based localization have focused on the description of local surface geometry, usually using pointclouds as the underlying representation. We propose a system for geometry-based localization that extracts features directly from an implicit surface representation: the Signed Distance Function (SDF). The SDF varies continuously through space, which allows the proposed system to extract and utilize features describing both surfaces and free-space. Through evaluations on public datasets, we demonstrate the flexibility of this approach, and show an increase in localization performance over state-of-the-art handcrafted surfaces-only descriptors. We achieve an average improvement of ~12% on an RGB-D dataset and ~18% on a LiDAR-based dataset. Finally, we demonstrate our system for localizing a LiDAR-equipped MAV within a previously built map of a search and rescue training ground.paper: arxiv.org/abs/2010.09378
Design and optimal control of a tiltrotor micro aerial vehicle for efficient omnidirectional flight aslteam 2020-08-07 | This video shows content related to our IJRR paper "Design and optimal control of a tiltrotor micro aerial vehicle for efficient omnidirectional flight".Link to paper: doi.org/10.1177/0278364920943654Abstract - Omnidirectional micro aerial vehicles are a growing field of research, with demonstrated advantages for aerial interaction and uninhibited observation. While systems with complete pose omnidirectionality and high hover efficiency have been developed independently, a robust system that combines the two has not been demonstrated to date. This paper presents the design and optimal control of a novel omnidirectional vehicle that can exert a wrench in any orientation while maintaining efficient flight configurations. The system design is motivated by the result of a morphology design optimization. A six degrees of freedom optimal controller is derived, with an actuator allocation approach that implements task prioritization, and is robust to singularities. Flight experiments demonstrate and verify the system's capabilities.Citation:@article{doi:10.1177/0278364920943654,author = {Mike Allenspach and Karen Bodie and Maximilian Brunner and Luca Rinsoz and Zachary Taylor and Mina Kamel and Roland Siegwart and Juan Nieto},title ={Design and optimal control of a tiltrotor micro-aerial vehicle for efficient omnidirectional flight},journal = {The International Journal of Robotics Research},pages = {0278364920943654},year = {2020},doi = {10.1177/0278364920943654},URL = {doi.org/10.1177/0278364920943654}}
Object Finding in Cluttered Scenes Using Interactive Perception aslteam 2020-06-26 | Object finding in clutter is a skill that requires perception of the environment and in many cases physical interaction. In robotics, interactive perception defines a set of algorithms that leverage actions to improve the perception of the environment, and vice versa use perception to guide the next action. Scene interactions are difficult to model, therefore, most of the current systems use predefined heuristics. This limits their ability to efficiently search for the target object in a complex environment. In order to remove heuristics and the need for explicit models of the interactions, in this work we propose a reinforcement learning based active and interactive perception system for scene exploration and object search. We evaluate our work both in simulated and in real-world experiments using a robotic manipulator equipped with an RGB and a depth camera, and compare our system to two baselines. The results indicate that our approach, trained in simulation only, transfers smoothly to reality and can solve the object finding task efficiently and with more than 88%success rate.This work was presented at ICRA 2020.arxiv.org/pdf/1911.07482.pdf
Voxgraph: Globally Consistent, Volumetric Mapping Using Signed Distance Function Submaps aslteam 2020-04-28 | Globally consistent dense maps are a key requirement for long-term robot navigation in complex environments. While previous works have addressed the challenges of dense mapping and global consistency, most require more computational resources than may be available on-board small robots. We propose a framework that creates globally consistent volumetric maps on a CPU and is lightweight enough to run on computationally constrained platforms. Our approach represents the environment as a collection of overlapping signed distance function (SDF) submaps and maintains global consistency by computing an optimal alignment of the submap collection. By exploiting the underlying SDF representation, we generate correspondence-free constraints between submap pairs that are computationally efficient enough to optimize the global problem each time a new submap is added. We deploy the proposed system on a hexacopter micro aerial vehicle (MAV) with an Intel i7-8650 U CPU in two realistic scenarios: mapping a large-scale area using a 3D LiDAR and mapping an industrial space using an RGB-D camera. In the large-scale outdoor experiments, the system optimizes a 120 × 80 m map in less than 4 s and produces absolute trajectory RMSEs of less than 1 m over 400 m trajectories. Our complete system, called voxgraph, is available as open source.Code: github.com/ethz-asl/voxgraphPaper: https://www.research-collection.ethz.ch/handle/20.500.11850/385682
Autonomous Fire Finding Using Aerial Robots aslteam 2020-04-13 | An autonomous mission to find a fire in a multi-story building. The mission is completed by a collaborating robotic team, consisting of a hexacopter and a tricopter. The approach exploits the mapping and precise control capabilities of each of the vehicles respectively.
Go Fetch: Mobile Manipulation in Unstructured Environments aslteam 2020-04-03 | With humankind facing new and increasingly large-scale challenges in the medical and domestic spheres, automation of the service sector carries a tremendous potential for improved efficiency, quality, and safety of operations. Mobile robotics can offer solutions with a high degree of mobility and dexterity, however these complex systems require a multitude of heterogeneous components to be carefully integrated into one consistent framework. This work presents a mobile manipulation system that combines perception, localization, navigation, motion planning and grasping skills into one common workflow for fetch and carry applications in unstructured indoor environments. The tight integration across the various modules is experimentally demonstrated on the task of finding a commonly-available object in an office environment, grasping it, and delivering it to a desired drop-off location.arXiv: arxiv.org/abs/2004.00899
Design and optimal control of a tiltrotor micro aerial vehicle for efficient omnidirectional flight aslteam 2020-04-01 | This video shows content related to our paper submission "Design and optimal control of a tiltrotor micro aerial vehicle for efficient omnidirectional flight".ArXiv preprint: arxiv.org/abs/2003.09512Abstract - Omnidirectional micro aerial vehicles are a growing field of research, with demonstrated advantages for aerial interaction and uninhibited observation. While systems with complete pose omnidirectionality and high hover efficiency have been developed independently, a robust system that combines the two has not been demonstrated to date. This paper presents the design and optimal control of a novel omnidirectional vehicle that can exert a wrench in any orientation while maintaining efficient flight configurations. The system design is motivated by the result of a morphology design optimization. A six degrees of freedom optimal controller is derived, with an actuator allocation approach that implements task prioritization, and is robust to singularities. Flight experiments demonstrate and verify the system's capabilities.
Active Interaction Force Control for Omnidirectional Aerial Contact-Based Inspection aslteam 2020-04-01 | This video presents supporting content for our paper submission titled "Active Interaction Force Control for Omnidirectional Aerial Contact-Based Inspection".ArXiv preprint: arxiv.org/abs/2003.09516Abstract - This paper presents and validates two approaches for active interaction force control and planning for omnidirectional aerial manipulation platforms, with the goal of aerial contact inspection in unstructured environments. We extend upon an axis-selective impedance controller to present a variable axis-selective impedance control which integrates direct force control for intentional interaction, using feedback from an on-board force sensor. The control approaches aim to reject disturbances in free flight, while handling unintentional interaction, and actively controlling desired interaction forces. A fully actuated and omnidirectional tilt-rotor aerial system is used to show capabilities of the control and planning methods. Experiments demonstrate disturbance rejection, push-and-slide interaction, and force controlled interaction in different flight orientations. The system is validated as a tool for non-destructive testing of concrete infrastructure, and statistical results of interaction control performance are presented and discussed.
