TUM - Institute for Cognitive Systems

Congratulations, Prof. Gordon Cheng, for becoming an RSJ Fellow this year! 🎉 Prof. Cheng was among the four fellows honored at the 42nd Annual Conference of The Robotic Society of Japan (RSJ). He was honored for his long-standing contributions to “science and technology”! 👏 Heartfelt congrats from the ICS team! 🙏

The next Doctoral Research Seminar next Monday will present 'Enjoying Time with Robots: Design Interactive Robots That Are Fun to Interact With' by prof. Albrecht Schmidt and Sven Mayer.🤖 July 1st, 2024, 10:30 am - 11:30 am in room 2026, Karlstr. 45. What does it take to make interacting with someone or something enjoyable? What qualities make us want to work with a particular person? When do we enjoy interacting with people, animals or things? Can we apply the answers to these questions to human-robot interaction? How can we design systems with higher levels of autonomy, such as smart environments, robots, and cars, so that we want to use them and interact with them? In particular, effective teaching of these systems is crucial to making them useful, enforcing desired behavior, and correcting undesired actions. In-situ learning and adaptation to specific contexts (e.g. culture, user type, preferences) are essential for effective real-world learning. However, motivating humans to teach machines is the key challenge. We argue that curiosity-driven behavior can lead to more natural interactions and increase human willingness to teach. We see this as a first step towards designing interactive robots that are fun to interact with. We hope to discuss with you our experiences from research and teaching in this area. #human-robot #interaction #robotics

📢The upcoming Doctoral Research Seminar this Monday will present "Learning while Sleeping: Integrating Sleep-Inspired Consolidation with Human Feedback Learning" by Imene Tarakli 🛌 🧠 🗓️June 24th 10:30-11:30 in room 2026, Karlstr. 45 Sleep plays a vital role in developmental learning. It allows the brain to consolidate daily learning experiences by replaying the memories accumulated throughout the day. In this work, we take inspiration from sleep and propose the Inverse Forward Offline Reinforcement Model (INFORM), a scalable framework that first learns a set of behaviours from human evaluative feedback, then consolidates the learning by applying an offline inverse reinforcement learning to the memorised trajectories. Experimental results demonstrate that INFORM is a feedback-efficient method that effectively learns an optimal policy that align with the intended behaviour of the human. A comparative analysis shows that the learnt policies are robust to dynamics changes in the environment and the recovered rewards allows the robot to be autonomous in its learning. #humanoid #robotics

🎉 Congratulations to our colleagues Fengyi Wang, Xiangyu Fu, prof. Nitish Thakor and prof. Gordon Cheng. Their paper "Object Classification Utilizing Neuromorphic Proprioceptive Signals in Active Exploration: Validated on a Soft Anthropomorphic Hand” has been accepted for publication for IEEE RAS EMBS 10th International Conference on Biomedical Robotics and Biomechatronics (BioRob 2024) taking place in Heidelberg (Germany) from 1-4 September 2024. 📅

College student Fengyi Wang just presented groundbreaking research on "Enhancing Roughness Classification Accuracy Using Neuromorphic Sensory Encoding and Spiking Neural Networks" at the Technical University of Munich Institute of Advanced Study event! This impressive work results from a successful collaboration between TUM - Institute for Cognitive Systems and John Hopkins University. Congratulations to Fengyi Wang and the team for their outstanding contribution to the field! #research #neuromorphic #spikingneuralnetworks #collaboration Nitish Thakor Mark Iskarous Gordon Cheng

The next Doctoral Research Seminar will present "SNAPP, An Agile Robotic Fish with 3-D Maneuverability for Open Water Swim " by Timothy Ng, HKU. 🤖 April 29th, 2024, 10:30 am - 11:30 am in room 2026, Karlstr. 45. Fish exhibit impressive locomotive performance and agility in complex underwater environments, using their undulating tails and pectoral fins for propulsion and maneuverability. Replicating these abilities in robotic fish is challenging; existing designs focus on either fast swimming or directional control at limited speeds, mainly within a confined environment. To address these limitations, we designed Snapp, an integrated robotic fish capable of swimming in open water with high speeds and full 3-dimensional maneuverability. A novel cyclic-differential method is layered on the mechanism. It integrates propulsion and yaw steering for fast course corrections. Two independent pectoral fins provide pitch and roll control. We evaluated Snapp in open water environments and demonstrated significant improvements in speed and maneuverability, achieving swimming speeds of 1.5 m/s (1.7 body lengths per second) and performing complex maneuvers, such as a figure-8 and S-shape trajectory. Instantaneous yaw changes of 15◦ in 0.4 s, a minimum turn radius of 0.85 m, and maximum pitch and roll rates of 3.5 and 1 rad/s, respectively, were recorded. Our results suggest that Snapp’s swimming capabilities have excellent practical prospects for open seas and contribute significantly to developing agile robotic fishes. #locomotive #roboticfish #innovation

The upcoming Doctoral Research Seminar next Monday will present "Bioinspired Fundamentals and Mechanisms for the Design of a Running Humanoid Robot“ by Konrad Fründ (German Aerospace Center (DLR))🤖 April 22nd, 2024, 10:30 – 11:30 in room 2026, Karlstr. 45 📅 The field of humanoid robots is currently experiencing significant growth, with most humanoids focusing on industrial tasks within walkable distances. However, the ability to rapidly traverse complex terrain might become one of the future tasks for humanoid robots, to solve their given tasks efficiently. Within the seminar we investigate the bioinspired fundamentals and mechanisms that drive the design of a running humanoid robot. We will explore the bioinspired principles from the remarkable locomotion capabilities observed in humans which can be translated into the design and control of humanoid robots. The fundamental biomechanical functions of the human running gait will be discussed and how they can serve as guidelines for innovative humanoid robot designs. A focus is then given to the kinematics of humanoid robots and their joint axes to investigate the effect of oblique axes on bipedal locomotion. The seminar talk will provide valuable insights into the interdisciplinary research field of biomechanics, human anatomy, robotics, and how it can push the boundaries of what humanoid robots can achieve. #humanoid #robotic #locomotion

Congratulations to Youssef Michel, Matteo Saveriano, and Dongheui Lee! The paper on "A Novel Safety-Aware Energy Tank Formulation Based on Control Barrier Functions " has been accepted for publication in the IEEE Robotics and Automation Letters (RA-L) 🎉 🔬 💪 In this work, they propose a novel formulation for energy tanks based on Control Barrier Functions (CBF). Their approach can simultaneously handle energy constraints to ensure passivity and enforce power limits in the system to enhance safety. Furthermore, their approach overcomes the discrete switching nature of classical energy tanks, ensuring smooth control commands. To achieve their desiderata, they formulate the tank as a second-order dynamical system, where they exploit CBF and Higher-Order CBF to obtain theoretical guarantees on fulfilling the energy and power constraints in the system. Furthermore, they derive conditions related to the tank design to ensure the passivity of the controlled robot. Their proposed approach is tested in a series of robot experiments where they validate their approach on tasks such as variable stiffness and force control, and in a scenario where it is desired to constrain the kinetic energy in the system. #Robotics #Technology #Innovation

🎉 Exciting News! Congrats to our Researchers Po-Chuan Chan, Natalia Paredes Acuña, Nicolas Berberich, Anna Polata, Gordon Cheng, for their Publication! 📚🔬 We are delighted to share the publication of our researchers' latest work titled "Feasibility Study of an Active Ball Joint Mechanism for a Wheelchair-Mounted 2-DOF Shoulder Exoskeleton" in ICAR! 🤝🤖 In this study, they evaluate the feasibility of using an active ball joint mechanism in a 3D-printed 2 degrees of freedom (DOF) exoskeleton for wheelchair users with upper-limb muscle weakness. The goal was to assess its effectiveness in assisting shoulder movements during activities of daily living (ADL). Results showed promising range of motion and reduced muscular fatigue, indicating the potential for a compact and accessible solution to support upper limbs in ADL tasks. On average, three healthy participants wearing the exoskeleton could reach 100% range of motion (ROM) of the required ADL workspace in shoulder flexion and 81% in shoulder abduction. Moreover, they had a delayed onset of muscular fatigue by 144s (113%) during muscular endurance tasks. During the exoskeleton's validation in two ADL tasks, the muscular activity in the anterior deltoid decreased by 48% during drinking tasks and 51% during combing tasks. These findings show that the proposed mechanism used in shoulder exoskeletons can provide a compact and accessible solution for assisting the upper limbs. Congrats to our exceptional researchers on this significant publication! 👏🌟 To explore the full paper, follow this link: https://lnkd.in/eTQbkxMc #exoskeleton #innovation #neuroengineering