Soft robots can revolutionize several applications with high demands on dexterity and safety. When operating these systems, real-time estimation and control require fast and accurate models. However, prediction with first-principles (FP) models is slow, and learned black-box models have poor generalizability. Physics-informed machine learning offers excellent advantages here, but it is currently limited to simple, often simulated systems without considering changes after training. We propose physics-informed neural networks (PINNs) for articulated soft robots (ASRs) with a focus on data efficiency. The amount of expensive real-world training data is reduced to a minimum — one dataset in one system domain. Two hours of data in different domains are used for a comparison against two gold-standard approaches: In contrast to a recurrent neural network, the PINN provides a high generalizability. The prediction speed of an accurate FP model is exceeded with the PINN by up to a factor of 467 at slightly reduced accuracy. This enables nonlinear model predictive control (MPC) of a pneumatic ASR. Accurate position tracking with the MPC running at 47 Hz is achieved in six dynamic experiments.

Fast and safe motion is crucial for the successful deployment of physically interactive robots. Parallel robots (PRs) offer the potential for higher speeds while maintaining the same energy limits due to their low moving masses. However, they require methods for contact detection and reaction while avoiding singularities and self-collisions. We address this issue and present SafePR - a unified approach for the detection and localization, including the distinction between collision and clamping to perform a reaction that is safe for humans and feasible for PRs. Our approach uses information from the encoders and motor currents to estimate forces via a generalized-momentum observer. Neural networks and particle filters classify and localize the contacts. We introduce reactions with redundancy resolution to avoid self-collisions and type-II singularities. Our approach detected and terminated 72 real-world collision and clamping contacts with end-effector speeds of up to 1.5m/s, each within 25-275ms. The forces were below the thresholds from ISO/TS 15066. By using built-in sensors, SafePR enables safe interaction with already assembled PRs without the need for new hardware components.

Sophisticated models can accurately describe deformations of continuum robots while being computationally demanding, which limits their application. Especially when considering sampling-based path planning, the model has to be evaluated frequently, which can lead to substantially increased computation times. We present a new approach to compute the entire shape of a tendon-driven continuum robot by a physics-informed neural network (PINN). The underlying physics is modelled with the Cosserat rod theory and incorporated into the PINN’s loss function. The boundary values for the training are obtained from a reference model, solved by the shooting method. Our approach allows for a computation of the learned Cosserat rod model multiple orders of magnitude faster than a publicly available reference model. The median position deviation from the reference model lies below 1mm (0.5% of the simulated robot length) for each of the robot’s 20 disks.

Using compliant materials leads to continuum robots undergoing large deformations. Their nonlinear behavior motivates the use of model-based controllers. They require state estimation as an essential step to be deployed. Available sensors are usually realized by introducing rigid bodies to the soft robot or inserting soft sensors made of materials different from the robot itself. Both approaches result in changes in the system’s dynamics. Optical measurements are problematic, especially in confined spaces. This can be avoided when the sensor is located at the robot's base. This paper studies the state estimation of a pneumatically actuated soft robot using the measured forces and torques at its base. For the first time, this is done using an unscented Kalman filter without restraining the dynamics to a planar or quasi-static motion while applying it to a real system. Real-time capability is achieved with our implementation. The state estimation is tested in a Cosserat rod simulation and on the physical system. The position is estimated with an accuracy of three to five millimeters for a 130 millimeter long pneumatic robot.

Snake robots offer considerable potential for endoscopic interventions due to their ability to follow curvilinear paths. Telemanipulation is an open problem due to hyper-redundancy, as input devices only allow a specification of six degrees of freedom. Our work addresses this by presenting a unified telemanipulation strategy which enables follow-the-leader locomotion and reorientation keeping the shape change as small as possible. The basis for this is a novel shape-fitting approach for solving the inverse kinematics in only a few milliseconds. Shape fitting is performed by maximizing the similarity of two curves using Fréchet distance while simultaneously specifying the position and orientation of the end effector. Telemanipulation performance is investigated in a study in which 14 participants controlled a simulated snake robot to locomote into the target area. In a final validation, pivot reorientation within the target area is addressed.

Parallel robots provide the potential to be lever-aged for human-robot collaboration (HRC) due to low collision energies even at high speeds resulting from their reduced moving masses. However, the risk of unintended contact with the leg chains increases compared to the structure of serial robots. As a first step towards HRC, contact cases on the whole parallel robot structure are investigated and a disturbance observer based on generalized momenta and measurements of motor current is applied. In addition, a Kalman filter and a second-order sliding-mode observer based on generalized momenta are compared in terms of error and detection time. Gearless direct drives with low friction improve external force estimation and enable low impedance. The experimental validation is performed with two force-torque sensors and a kinetostatic model. This allows a new identification method of the motor torque constant of an assembled parallel robot to estimate external forces from the motor current and via a dynamics model. A Cartesian impedance control scheme for compliant robot-environmental dynamics with stiffness from 0.1-2N/mm and the force observation for low forces over the entire structure are validated. The observers are used for collisions and clamping at velocities of 0.4-0.9 m/s for detection within 9–58 ms and a reaction in the form of a zero-g mode.

This paper presents a new approach to determine single and multiple simultaneous contact forces on a tendon-driven continuum robot (CR). The estimation is based solely on the proprioceptive tendon force and length sensors that are already present. Unlike for rigid-body robots, only indirect measurements of the external forces' deflection is available. The required full kinetostatic model, which is prone to local minima due to the unknown contacts, is solved with a particle filter. The method is validated by simulative studies and experimental investigations on a new robot setup for visual inspection of aircraft engines. The algorithm allows the estimation of single contacts with an error up to 4.43 mm or 2.9 % of the robot's length. Multiple contacts can only be correctly determined at the two distal of the three segments.

Parallel robots (PRs) offer the potential for safe human-robot collaboration because of their low moving masses. Due to the in-parallel kinematic chains, the risk of contact in the form of collisions and clamping at a chain increases. Ensuring safety is investigated in this work through various contact reactions on a real planar PR. External forces are estimated based on proprioceptive information and a dynamics model, which allows contact detection. Retraction along the direction of the estimated line of action provides an instantaneous response to limit the occurring contact forces within the experiment to 70 N at a maximum velocity of 0.4 m/s. A reduction in the stiffness of a Cartesian impedance control is investigated as a further strategy. For clamping, a feedforward neural network (FNN) is trained and tested in different joint angle configurations to classify whether a collision or clamping occurs with an accuracy of 80%. A second FNN classifies the clamping kinematic chain to enable a subsequent kinematic projection of the clamping joint angle onto the rotational platform coordinates. In this way, a structure opening is performed in addition to the softer retraction movement. The reaction strategies are compared in real-world experiments at different velocities and controller stiffnesses to demonstrate their effectiveness. The results show that in all collision and clamping experiments the PR terminates the contact in less than 130 ms.

Parallel robots (PRs) allow for higher speeds in human-robot collaboration due to their lower moving masses but are more prone to unintended contact. For a safe reaction, knowledge of the location and force of a collision is useful. A novel algorithm for collision isolation and identification with proprioceptive information for a real PR is the scope of this work. To classify the collided body, the effects of contact forces at the links and platform of the PR are analyzed using a kinetostatic projection. This insight enables the derivation of features from the line of action of the estimated external force. The significance of these features is confirmed in experiments for various load cases. A feedforward neural network (FNN) classifies the collided body based on these physically modeled features. Generalization with the FNN to 300k load cases on the whole robot structure in other joint angle configurations is successfully performed with a collision-body classification accuracy of 84% in the experiments. Platform collisions are isolated and identified with an explicit solution, while a particle filter estimates the location and force of a contact on a kinematic chain. Updating the particle filter with estimated external joint torques leads to an isolation error of less than 3 cm and an identification error of 4 N in a real-world experiment.

This work presents a non-parametric spatiotemporal model for mapping human activity by mobile autonomous robots in a long-term context. Based on Variational Gaussian Process Regression, the model incorporates prior information of spatial and temporal-periodic dependencies to create a continuous representation of human occurrences. The inhomogeneous data distribution resulting from movements of the robot is included in the model via a heteroscedastic likelihood function and can be accounted for as predictive uncertainty. Using a sparse formulation, data sets over multiple weeks and several hundred square meters can be used for model creation. The experimental evaluation, based on multi-week data sets, demonstrates that the proposed approach outperforms the state of the art both in terms of predictive quality and subsequent path planning.

Robot manipulators are often deployed in tool-symmetric tasks, which only requires defining end effector position and pointing direction. In this case six-axis serial industrial robots and full-mobility (spatial) parallel robots have one degree of task redundancy. Using Cardan angles as orientation coordinates, a unified formulation of the position-level and second-order inverse kinematics problem is set up for both robot types. An efficient scheme for difference-quotient approximation of gradients of performance criteria for projection into the task redundancy's nullspace is presented. The simulation example of a hexapod robot shows that avoiding and exiting parallel robot singularities of type II is possible with the nullspace of all joints. The nullspace controller scheme can be used in offline trajectory optimization and in online motion generation.

This paper presents an extension of the dual-layer model predictive control (MPC) for industrial DC microgrids. This structure allows an easier integration of stochastic methods on all levels. Moreover, in the new three-layer MPC, energy storages are proposed for each level. As a proof of concept, a control is presented for the new layer, the primary control level in microgrids. The MPC as a central control strategy, has the disadvantage that a new model has to be created for new bus participants. By a simply expandable state space model and a parameterization with data sheet values, this disadvantage can be refuted. Further, for the application on a programmable logic controller (PLC), a simpler and faster MPC with an equal control performance, is presented.

Soft material robotics is a rather young research field in the robotics and material science communities. A popular design is the soft pneumatic actuator (SPA) which, if connected serially, becomes a highly compliant manipulator. This high compliance makes it possible to adapt to the environment and in the future might be very useful for manipulation tasks in narrow and wound environments. A central topic is the modelling of the manipulators. While comparatively rigid continuum robots are build of metal or other materials, that conduct a linear behaviour, the material used in soft material robotics often exhibits a nonlinear stress-strain relationship. In this paper we contribute an identification method for material parameters and data-based approach within the constitutive equations of a Cosserat rod model. We target bending and extension stiffness, consider shear and neglect torsional strains. The proposed method is applicable to any continuum robot which can be modelled by the classic theory of special Cosserat rods, including constraint models, and shows great improvement in experimental results with mean position errors of 0.59% reference length.

In this paper the first robot-based prototype of a semi-autonomous upper-limb exoprosthesis is introduced, unifying exoskeletons and prostheses [1]. A central goal of this work is to minimize unnecessary interaction forces on the residual limb by compensating gravity effects via a upper body grounded exoskeleton. Furthermore, the exoskeleton provides the residual limb's kinematic data that allows to design more intelligent coordinated control concepts. The soft-robotics design of a prototype consisting of a transhumeral prosthesis and a robot-based exoskeleton substitute is outlined. For this class of hybrid systems a human embodied dynamics model and semi-autonomous coordinated motion strategies are derived. Here, in contrast to established standard sequential strategies all joints are moved simultaneously according to a desired task. In combination with an app-based programming framework the strategy goals are set either user-based via kinesthetic teaching or autonomously via 3D visual perception. This enables the user to execute tasks faster and more intuitive. First experimental evaluations show promising performance with a healthy subject.

High-performance collision handling, which is divided into the five phases detection, isolation, estimation, classification and reaction, is a fundamental robot capability for safe and sensitive operation/interaction in unknown environments. For complex humanoid robots collision handling is obviously significantly more complex than for classical static manipulators. In particular, the robot stability during the collision reaction phase has to be carefully designed and relies on high fidelity contact information that is generated during the first three phases. In this paper, a unified realtime algorithm is presented for determining unknown contact forces and contact locations for humanoid robots based on proprioceptive sensing only, i.e. joint position, velocity and torque, as well as force/torque sensing along the structure. The proposed scheme is based on nonlinear model-based momentum observers that are able to recover the unknown contact forces and the respective locations. The dynamic loads acting on internal force/torque sensors are also corrected based on a novel nonlinear compensator. The theoretical capabilities of the presented methods are evaluated in simulation with the Atlas robot. In summary, we propose a full solution to the problem of collision detection, collision isolation and collision identification for the general class of humanoid robots.

Soft robotics methods such as impedance control and reflexive collision handling have proven to be a valuable tool to robots acting in partially unknown and potentially unstructured environments. Mainly, the schemes were developed with focus on classical electromechanically driven, torque controlled robots. There, joint friction, mostly coming from high gearing, is typically decoupled from link-side control via suitable rigid or elastic joint torque feedback. Extending and applying these algorithms to stiff hydraulically actuated robots poses problems regarding the strong influence of friction on joint torque estimation from pressure sensing, i.e. link-side friction is typically significantly higher than in electromechanical soft robots. In order to improve the performance of such systems, we apply state-of-the-art fault detection and estimation methods together with observer-based disturbance compensation control to the humanoid robot Atlas. With this it is possible to achieve higher tracking accuracy despite facing significant modeling errors. Compliant end-effector behavior can also be ensured by including an additional force/torque sensor into the generalized momentum-based disturbance observer algorithm from [1].