Stefano Mintchev

Last Name

Mintchev

First Name

Stefano

ORCID

0000-0001-6272-0212

Organisational unit

09718 - Mintchev, Stefano / Mintchev, Stefano

Show all metadata

Search Results

Publications 1 - 10 of 34

Learning Occluded Branch Depth Maps in Forest Environments Using RGB-D Images
Geckeler, Christian; Aucone, Emanuele; Schnider, Yannick; et al. (2024)
IEEE Robotics and Automation Letters
Covering over a third of all terrestrial land area, forests are crucial environments; as ecosystems, for farming, and for human leisure. However, they are challenging to access for environmental monitoring, for agricultural uses, and for search and rescue applications. To enter, aerial robots need to fly through dense vegetation, where foliage can be pushed aside, but occluded branches pose critical obstacles. Therefore, we propose pixel-wise depth regression of occluded branches using three different U-Net inspired architectures. Given RGB-D input of trees with partially occluded branches, the models estimate depth values of only the wooden parts of the tree. A large photorealistic simulation dataset comprising around 44 K images of nine different tree species is generated, on which the models are trained. Extensive evaluation and analysis of the models on this dataset is shown. To improve network generalization to real-world data, different data augmentation and transformation techniques are performed. The approaches are then also successfully demonstrated on real-world data of broadleaf trees from Swiss temperate forests and a tropical Masoala Rainforest. This work showcases the previously unexplored task of frame-by-frame pixel-based occluded branch depth reconstruction to facilitate robot traversal of forest environments.
Whisker-based Haptic Perception System for Branch Detection in Dense Vegetation
Andresen, Leiv; Aucone, Emanuele; Mintchev, Stefano (2022)
2022 IEEE 5th International Conference on Soft Robotics (RoboSoft)
Dense vegetation is an example of a structurally complex environment that robots encounter in many outdoor applications such as agriculture, environmental monitoring, or search and rescue. For a robot to safely navigate or perform manipulation tasks in dense vegetation, it is important to detect and distinguish rigid branches from the surrounding soft foliage. This task is challenging for traditional sensing approaches, such as vision, because the foliage can partially or totally occlude the branches. We present a haptic sensing system capable of detecting contact with the vegetation and locating branches occluded by soft foliage. The system follows a vision-based tactile sensing approach consisting of an array of compliant whiskers, with fiducial markers attached to them, and a camera to track their displacement. When the whiskers are inserted in the vegetation and the array is oscillated, rigid branches cause the whiskers to deflect significantly, while deflections caused by softer foliage are smaller. We developed a sampling strategy and a software pipeline to detect the location of the branches based on the deflection magnitude of the whiskers. Upon indoor and outdoor experiments with artificial and natural vegetation, we demonstrate the ability to locate branches among compliant foliage.
Optical Tactile Sensing for Aerial Multi-Contact Interaction: Design, Integration, and Evaluation
Aucone, Emanuele; Sferrazza, Carmelo; Gregor, Manuel; et al. (2025)
IEEE Transactions on Robotics
Distributed tactile sensing for multi-force detection is crucial for various aerial robot interaction tasks. However, current contact sensing solutions on drones only exploit single end-effector sensors and cannot provide distributed multi-contact sensing. Designed to be easily mounted at the bottom of a drone, we propose an optical tactile sensor that features a large and curved soft sensing surface, a hollow structure and a new illumination system. Even when spaced only 2cm apart, multiple contacts can be detected simultaneously using our software pipeline, which provides real-world quantities of 3D contact locations (mm) and 3D force vectors (N), with an accuracy of 1.5 mm and 0.17 N respectively. We demonstrate the sensor's applicability and reliability onboard and in real-time with two demos related to i) the estimation of the compliance of different perches and subsequent re-alignment and landing on the stiffer one, and ii) the mapping of sparse obstacles. The implementation of our distributed tactile sensor represents a significant step towards attaining the full potential of drones as versatile robots capable of interacting with and navigating within complex environments.
Encoding mechanical intelligence using ultra-programmable joints
Wu, Rui; Girardi, Luca; Mintchev, Stefano (2025)
Science Advances
Animal bodies act as physical controllers, with their finely-tuned passive me chanical responses physically “encoding” complex movements and environmental interactions. This capability allows animals to perform challenging tasks with minimal muscular or neural activities, a phenomenon known as Embodied In telligence (EI). EI offers promising prospects for robots with enhanced agility, efficiency, and simplified control systems. However, designing robotics EI remains significantly constrained due to the lack of physical bodies with abundant tune able parameters—such as tuneable stiffness—which is a critical factor akin to the programmable parameters of a neural network. Therefore, we introduce Elastic Rolling Cam (ERC), a universally programmable embodiment with accurately inverse-designable rotational stiffness. We have shown in simulations that the ERC can closely replicate 100,000 randomly-generated stiffness profiles. Proto types ranging from millimeters to centimeters were manufactured to validate the programmability and scalability, as well as an impact-tolerance that resem bles bone-tendon joints. To illustrate that EI can be encoded by programming the ERC’s stiffness response, we designed a bipedal robot with optimized ERC 1 passive knees, achieving energy-efficient, open-loop stable walking across uneven terrain using simple harmonic hip actuation. We also demonstrated a quadcopter drone with ERC elbow joints encoding an impact-activated, dual-state morphing, enabling it to fly through narrow passages without additional morphing actuators.
Design and Control of a Micro Overactuated Aerial Robot with an Origami Delta Manipulator
Cuniato, Eugenio; Geckeler, Christian; Brunner, Maximilian; et al. (2023)
2023 IEEE International Conference on Robotics and Automation (ICRA)
This work presents the mechanical design and control of a novel small-size and lightweight Micro Aerial Vehicle (MAV) for aerial manipulation. To our knowledge, with a total take-off mass of only 2.0 kg, the proposed system is the most lightweight Aerial Manipulator (AM) that has 8-DOF independently controllable: 5 for the aerial platform and 3 for the articulated arm. We designed the robot to be fully-actuated in the body forward direction. This allows independent pitching and instantaneous force generation, improving the platform's performance during physical interaction. The robotic arm is an origami delta manipulator driven by three servomotors, enabling active motion compensation at the end-effector. Its composite multimaterial links help reduce the weight, while their flexibility allow for compliant aerial interaction with the environment. In particular, the arm's stiffness can be changed according to its configuration. We provide an in depth discussion of the system design and characterize the stiffness of the delta arm. A control architecture to deal with the platform's overactuation while exploiting the delta arm is presented. Its capabilities are experimentally illustrated both in free flight and physical interaction, highlighting advantages and disadvantages of the origami's folding mechanism.
HEDGEHOG: Drone Perching on Tree Branches with High-Friction Origami Spines
Kirchgeorg, Steffen; Mintchev, Stefano (2022)
IEEE Robotics and Automation Letters
The collection of environmental and biodiversity data is essential to manage, preserve and restore forests, but this task remains challenging due to the inaccessibility of these ecosystems. Compared to human intervention, aerial robots can access tree canopies, but their limited flight time and noise continue to stall widespread application. To address this challenge, we present a perching mechanism which allows small drones to rest on overhanging branches and extend their mission while remaining silent. We developed an origami spine with two folding flaps containing a layer of high-friction material. When the spine engages with a branch, the flaps open and conform to irregular branch surfaces generating sufficient friction to support the weight of a drone. With HEDGEHOG, a drone integrating multiple spines on a protective cage, we demonstrated its application in a controlled indoor as well as in a forest environment. We modelled the perching strategy and measured the effects of materials and geometric parameters on the drone's perching performance. By leveraging interactions with nature, our drone can perch on tree branches with diameters up to 86 mm and inclined up to +-15 and potentially remain in the canopy for extended periods of time to acquire data or monitor returning wildlife.
eProbe: Sampling of Environmental DNA within Tree Canopies with Drones
Kirchgeorg, Steffen; Chang, Jia Jin Marc; Ip, Yin Cheong Aden; et al. (2024)
Environmental Science & Technology
Environmental DNA (eDNA) analysis is a powerful tool for studying biodiversity in forests and tree canopies. However, collecting representative eDNA samples from these high and complex environments remains challenging. Traditional methods like surface swabbing or tree rolling are labor-intensive and require significant effort to achieve adequate coverage. This study proposes a novel approach for unmanned aerial vehicles (UAVs) to collect eDNA within tree canopies using a surface swabbing technique. The method involves lowering a probe from a hovering UAV into the canopy, collecting eDNA as it descends and ascends through branches and leaves. To achieve this, a custom-designed robotic system was developed, featuring a winch and a probe for eDNA collection. The design of the probe was optimized and a control logic for the winch developed to reduce the risk of entanglement while ensuring sufficient interaction force to facilitate transfer of eDNA onto the probe. The effectiveness of this method was demonstrated during the XPRIZE Rainforest Semi-Finals as ten eDNA samples were collected from the rainforest canopy and a total of 152 molecular operational taxonomic units (MOTUs) were identified using eDNA metabarcoding. We further investigate how the number of probe interactions with vegetation, the penetration depth, and the sampling duration influence the DNA concentration and community composition of the samples.
The current state and future outlook of rescue robotics
Delmerico, Jeffrey; Mintchev, Stefano; Giusti, Alessandro; et al. (2019)
Journal of Field Robotics
Synergistic morphology and feedback control for traversal of unknown compliant obstacles with aerial robots
Aucone, Emanuele; Geckeler, Christian; Morra, Daniele; et al. (2024)
Nature Communications
Animals traverse vegetation by direct physical interaction using their entire body to push aside and slide along compliant obstacles. Current drones lack this interaction versatility that stems from synergies between body morphology and feedback control modulated by sensing. Taking inspiration from nature, we show that a task-oriented design allows a drone with a minimalistic controller to traverse obstacles with unknown elastic responses. A discoid sensorized shell allows to establish and sense contacts anywhere along the shell and facilitates sliding along obstacles. This simplifies the formalization of the control strategy, which does not require a model of the interaction with the environment, nor high-level switching conditions for alternating between pushing and sliding. We utilize an optimization-based controller that ensures safety constraints on the robot’s state and dampens the oscillations of the environment during interaction, even if the elastic response is unknown and variable. Experimental evaluation, using a hinged surface with three different stiffness values ranging from 18 to 155.5 N mm rad−1, validates the proposed embodied aerial physical interaction strategy. By also showcasing the traversal of isolated branches, this work makes an initial contribution toward enabling drone flight across cluttered vegetation, with potential applications in environmental monitoring, precision agriculture, and search and rescue.
Design, Modeling and Control of AVOCADO: A Multimodal Aerial-Tethered Robot for Tree Canopy Exploration
Kirchgeorg, Steffen; Aucone, Emanuele; Wenk, Florian; et al. (2024)
IEEE Transactions on Robotics
Forests provide vital resources and services for humanity, but preserving and restoring them is challenging due to the difficulty of obtaining actionable data, especially in inaccessible areas such as forest canopies. To address this, we follow the lead of arboreal animals that exploit multiple modes of locomotion.We combine aerial and tethered movements to enable AVOCADO to navigate with in a tree canopy.Starting from the top of a tree, it can descend with the tether and maneuver around obstacles with thrusters. We extend our previous work with a new mechanical design with a protective shell, increased computational power and cameras for state estimation. We introduce a dynamic model and simulation, and perform a quasi-static and dynamic validation. For autonomy, we derive a control framework in simulation to regulate tether length,tilt and heading, before transfer to the robot. We evaluate the controllers for trajectory tracking through experiments. AVOCADO can follow trajectories around obstacles and reject disturbances on the tether. Exploiting multimodal mobility will advance the exploration of tree canopies to actively monitor the true value of our forests.

Publications 1 - 10 of 34

Stefano Mintchev

Last Name

First Name

ORCID

Organisational unit

Filters

Search Results