Soft robots offer more flexibility, compliance, and adaptability than traditional rigid robots. They are also typically lighter and cheaper to manufacture. However, their use in real-world applications is limited due to modeling challenges and difficulties in integrating effective proprioceptive sensors. Large-scale soft robots (≈ two meters in length) have greater modeling complexity due to increased inertia and related effects of gravity. Common efforts to ease these modeling difficulties such as assuming simple kinematic and dynamics models also limit the general capabilities of soft robots and are not applicable in tasks requiring fast, dynamic motion like throwing and hammering. To overcome these challenges, we propose a data-efficient Bayesian optimization-based approach for learning control policies for dynamic tasks on a large-scale soft robot. Our approach optimizes the task objective function directly from commanded pressures, without requiring approximate kinematics or dynamics as an intermediate step. We demonstrate the effectiveness of our approach through both simulated and real-world experiments.

Despite the existence of robots that can physically lift heavy loads, robots that can collaborate with people to move heavy objects are not readily available. This article makes progress toward effective human-robot co-manipulation by studying 30 human-human dyads that collaboratively manipulated an object weighing \(27 \mathrm{kg}\) without being co-located (i.e., participants were at either end of the extended object). Participants maneuvered around different obstacles with the object while exhibiting one of four modi–the manner or objective with which a team moves an object together–at any given time. Using force and motion signals to classify modus or behavior was the primary objective of this work. Our results showed that two of the originally proposed modi were very similar, such that one could effectively be removed while still spanning the space of common behaviors during our co-manipulation tasks. The three modi used in classification were quickly , smoothly and avoiding obstacles . Using a deep convolutional neural network (CNN), we classified three modi with up to 89% accuracy from a validation set. The capability to detect or classify modus during co-manipulation has the potential to greatly improve human-robot performance by helping to define appropriate robot behavior or controller parameters depending on the objective or modus of the team.

In this paper we present a novel kinematic representation of a soft continuum robot to enable full shape estimation using a purely geometric solution. The kinematic representation involves using length varying piecewise constant curvature segments to describe the deformed shape of the robot. Based on this kinematic representation, we can use overlapping length sensors to estimate the shape of continuously deformable bodies without prior knowledge of the current loading conditions. We show an implementation that assumes one change in curvature along the length of a joint, using string potentiometers as an arc length sensor, and an orientation measurement from the tip of the continuum joint. For 56 randomized joint configurations, we estimate the shape of a 250 mm long continually deformable robot with less then 2.5 mm of average error. The average error is reported for each of the 10 different equally spaced points along the length, demonstrating the ability to accurately represent the full shape of the soft robot.

In this paper, we propose a new tractable ordinary differential equation formulation for dynamic simulation of fabric- reinforced inflatable soft robots. The method performs a lumped-parameter discretization of the continuum robot into discrete discs (inertia), spring elements, and threads (representing the inextensible fabric reinforcement). Using the repetition in the structure of the Lagrangian formulation of the dynamic equations of motion, a method is developed that outputs machine- readable analytical expressions for the equations of motion. The method does not require symbolic computation of derivatives. The recursive nature allows us to scale the model to an arbitrary number$N$discs, and can represent buckling, twisting, and pleating that is commonly seen in very soft robots. The expressions generated were validated against manually-derived equations of motion for the two-disc case using both Lagrangian and Newton-Euler means. A simulation environment which parses and evaluates the analytical expressions generated at run-time was used to numerically integrate and predict the response of a four-disc example robot. Trajectories observed varied smoothly and plausibly predicted the behavior envisioned in robots like these.

In this paper we present a novel approach to accomplishing soft robot configuration estimation and control using RGB-D cameras and SLAM-based methods. By placing cameras on the unactuated sections of our large-scale (approximately 2 meters long) pneumatic soft robot, we can map an environment and then estimate the orientation of the robot links using landmark-based localization. Using the orientations of each camera we can solve for the joint configurations between them. We first show that this method works for a traditional rigid robot (Baxter) where we can compare against the ground truth encoder values. For Baxter, the median joint angle error was on the order of 1-2◦. We then show that the SLAM-based method provides estimates for soft robot configuration that are within 1◦when compared to our past methods of using a HTC Vive Tracker. While HTC Vive Trackers and commonly used motion capture systems require externally mounted sensors placed in the robot’s environment, the SLAM-based estimation method presented here works in any visually feature-rich environment. Finally we show that this method of estimation is effective for closed-loop control of soft robots by controlling our large-scale soft robot through a series of joint configurations.

We present a method for the design optimization of a soft, inflatable robot. The method described utilizes a multi-objective fitness function together with custom, platform-specific metrics related to the dexterity and load-bearing capacity of inflatable manipulators. Candidate designs are scored by computing these metrics at many randomly generated configurations and then by appropriately combining these scores within the multi-objective optimization framework. High performing designs are propagated through a genetic algorithm. The final result is a set of diverse, optimal designs lying along a Pareto front spanning the design space. By examining variations and trade-offs within this set, a designer can more appropriately choose design parameters for a target application. This is especially relevant for robots with many design parameters that can quickly be manufactured as is the case with emerging, soft robot technologies.

A novel algorithm for planning robotic manipulation tasks is presented in which the base position and joint motions of a robot are simultaneously optimized to follow a smooth desired end-effector trajectory. During the optimization routine, the manipulator's base position and joint motions are planned simultaneously by strategically moving a set of virtual robot arms (each representing a single configuration in a sequence) branching from a common base to a number of assigned target poses associated with a task. Additional goals (e.g. collision avoidance) and hard constraints, including joint limits are also incorporated. The optimization problem at the core of this method is a quadratic program, allowing constrained high-dimensional problems to be solved in very little time. This method has successfully planned motions allowing an 8-DOF manipulator to paint walls, and has proven to be highly efficient and scalable in practice.

Although inflatable soft robots are not yet a common robot platform, air leaking from the internal structure is a common and undesirable mode of failure for these platforms. In this paper we demonstrate a method to detect leaks in the structural chamber of an inflatable, pneumatically actuated robot. We then show that our method can adaptively lower commanded joint stiffness which slows the mass flow rate of the leak. This extends the operational life of the robot by decreasing long term error during operation by as much as 50% of the steady state error at the end effector when compared to the same leak if our adaptation method is not used. In future applications where we expect soft, inflatable robots to be useful, our methods can enable failure mitigation in resource-limited situations such as space exploration or disaster response.

Soft robot research has led to the development of platforms that should allow for better performance when working in uncertain or dynamic environments. The potential improvement in performance of these platforms ranges from mechanical robustness to high forces, to applying lower incidental contact forces in uncertain situations. However, the promise of these platforms is limited by the difficulty of controlling them. In this paper, we present preliminary results on simultaneously controlling stiffness and position for a pneumatically actuated soft robot. Improving on our prior work, we show that by including the pressure in our soft robot actuation chambers as state variables we can improve our average rise time by up to 137%, settling time by 119%, and overshoot by 853%. In addition to these improvements, we can now control both joint position and stiffness simultaneously. This performance improvement comes from using Model Predictive Control running at 300 Hz with improved dynamic models of the soft robot. High performance control of soft robot joints, such as the joint presented in this paper, will enable a wide range of robot applications that were previously difficult or impossible due to the rigid nature of traditional robot linkages and actuation schemes.

Biological organisms benefit from tactile sensing across the entire surfaces of their bodies. Robots may also be able to benefit from this type of sensing, but fully covering a robot with robust and capable tactile sensors entails numerous challenges. To date, most tactile sensors for robots have been used to cover rigid surfaces. In this paper, we focus on the challenge of tactile sensing across articulated joints, which requires sensing across a surface whose geometry varies over time. We first demonstrate the importance of sensing across joints by simulating a planar arm reaching in clutter and finding the frequency of contact at the joints. We then present a simple model of how much a tactile sensor would need to stretch in order to cover a 2 degree-of-freedom (DoF) wrist joint. Next, we describe and characterize a new tactile sensor made with stretchable fabrics. Finally, we present results for a stretchable sleeve with 25 tactile sensors that covers the forearm, 2 DoF wrist, and end effector of a humanoid robot. This sleeve enabled the robot to reach a target in instrumented clutter and reduce contact forces.

An omnidirectional Mecanum base allows for more flexible mobile manipulation. However, slipping of the Mecanum wheels results in poor dead-reckoning estimates from wheel encoders, limiting the accuracy and overall utility of this type of base. We present a system with a downward-facing camera and light ring to provide robust visual odometry estimates. We mounted the system under the robot which allows it to operate in conditions such as large crowds or low ambient lighting. We demonstrate that the visual odometry estimates are sufficient to generate closed-loop PID (Proportional Integral Derivative) and LQR (Linear Quadratic Regulator) controllers for motion control in three different scenarios: waypoint tracking, small disturbance rejection, and sideways motion. We report quantitative measurements that demonstrate superior control performance when using visual odometry compared to wheel encoders. Finally, we show that this system provides high-fidelity odometry estimates and is able to compensate for wheel slip on a four-wheeled omnidirectional mobile robot base.