This paper provides a system overview about ANYmal, a quadrupedal robot developed for operation in harsh environments. The 30 kg, 0.5 m tall robotic dog was built in a modular way for simple maintenance and user-friendly handling, while focusing on high mobility and dynamic motion capability. The system is tightly sealed to reach IP67 standard and protected to survive falls. Rotating lidar sensors in the front and back are used for localization and terrain mapping and compact force sensors in the feet provide accurate measurements about the contact situations. The variable payload, such as a modular pan-tilt head with a variety of inspection sensors, can be exchanged depending on the application. Thanks to novel, compliant joint modules with integrated electronics, ANYmal is precisely torque controllable and very robust against impulsive loads during running or jumping. In a series experiments we demonstrate that ANYmal can execute various climbing maneuvers, walking gaits, as well as a dynamic trot and jump. As special feature, the joints can be fully rotated to switch between X-and O-type kinematic configurations. Detailed measurements unveil a low energy consumption of 280 W during locomotion, which results in an autonomy of more than 2 h.

This paper provides insight into the application of the quadrupedal robot ANYmal in outdoor missions of industrial inspection (autonomous robot for gas and oil sites[ARGOS] challenge) and search and rescue (European Robotics League (ERL) Emergency Robots). In both competitions, the legged robot had to autonomously and semiautonomously navigate in real-world scenarios to complete high-level tasks such as inspection and payload delivery. In the ARGOS competition, ANYmal used a rotating light detection and ranging sensor to localize on the industrial site and map the terrain and obstacles around the robot. In the ERL competition, additional realtime kinematic-global positioning system was used to colocalize the legged robot with respect to a micro aerial vehicle that creates maps from the aerial view. The high mobility of legged robots allows overcoming large obstacles, for example, steps and stairs, with statically and dynamically stable gaits. Moreover, the versatile machine can adapt its posture for inspection and payload delivery. The paper concludes with insight into the general learnings from the ARGOS and ERL challenges.

Reduction of the system complexity is currently one of the main challenges for efficient and versatile legged robot locomotion. In this paper, we present a new one legged hopping robot called CHIARO, which is equipped with a curved foot. Even though the robot has no sensory feedback and consists of only two rigid bodies and one spring loaded joint with parallel actuation, it is able to achieve stable forwardhopping over a wide range of parameters and forward-speeds. Operating at natural hopping frequency, the parallel actuation shows good efficiency. This paper presents an approach to determine stability and efficiency of a highly non-linear mechanical system. By implementing a two dimensional numerical model, taking into account ground contact forces by a Newtonian kinematic impact-and coulomb friction law, we conducted a thorough parameter analysis based on a series of simulations. The comparison of the simulation and real world experiments shows good accordance, which qualifies the simulation for parameter optimization including prediction of robot stability and efficiency.

Scientific visualization developed successful methods for scalar and vector fields. For tensor fields, however, effective, interactive visualizations are still missing despite progress over the last decades. We present a general approach for the generation of separating surfaces in symmetric, second-order, three-dimensional tensor fields. These surfaces are defined as fiber surfaces of the invariant space, i.e. as pre-images of surfaces in the range of a complete set of invariants. This approach leads to a generalization of the fiber surface algorithm by Klacansky et al. [16] to three dimensions in the range. This is due to the fact that the invariant space is three-dimensional for symmetric second-order tensors over a spatial domain. We present an algorithm for surface construction for simplicial grids in the domain and simplicial surfaces in the invariant space. We demonstrate our approach by applying it to stress fields from component design in mechanical engineering.