In this work we propose a learning-based approach to box loco-manipulation for a humanoid robot. This is a particularly challenging problem due to the need for whole-body coordination in order to lift boxes of varying weight, position, and orientation while maintaining balance. To address this challenge, we present a sim-to-real reinforcement learning approach for training general box pickup and carrying skills for the bipedal robot Digit. Our reward functions are designed to produce the desired interactions with the box while also valuing balance and gait quality. We combine the learned skills into a full system for box loco-manipulation to achieve the task of moving boxes from one table to another with a variety of sizes, weights, and initial configurations. In addition to quantitative simulation results, we demonstrate successful sim-to-real transfer on the humanoid robot Digit. To our knowledge this is the first demonstration of a learned controller for such a task on real world hardware.

Reinforcement learning (RL) for bipedal locomotion has recently demonstrated robust gaits over moderate terrains using only proprioceptive sensing. However, such blind controllers will fail in environments where robots must anticipate and adapt to local terrain, which requires visual perception. In this paper, we propose a fully-learned system that allows bipedal robots to react to local terrain while maintaining commanded travel speed and direction. Our approach first trains a controller in simulation using a heightmap expressed in the robot’s local frame. Next, data is collected in simulation to train a heightmap predictor, whose input is the history of depth images and robot states. We demonstrate that with appropriate domain randomization, this approach allows for successful sim-to-real transfer with no explicit pose estimation and no fine-tuning using real-world data. To the best of our knowledge, this is the first example of sim-to-real learning for vision-based bipedal locomotion over challenging terrains.

A necessary capability for humanoid robots is the ability to stand and walk while rejecting natural disturbances. Recent progress has been made using sim-to-real reinforcement learning (RL) to train such locomotion controllers, with approaches differing mainly in their reward functions. However, prior works lack a clear method to systematically test new reward functions and compare controller performance through repeatable experiments. This limits our understanding of the trade-offs between approaches and hinders progress. To address this, we propose a low-cost, quantitative benchmarking method to evaluate and compare the real-world performance of standing and walking (SaW) controllers on metrics like command following, disturbance recovery, and energy efficiency. We also revisit reward function design and construct a minimally constraining reward function to train SaW controllers. We experimentally verify that our benchmarking framework can identify areas for improvement, which can be systematically addressed to enhance the policies. We also compare our new controller to state-of-the-art controllers on the Digit humanoid robot. The results provide clear quantitative trade-offs among the controllers and suggest directions for future improvements to the reward functions and expansion of the benchmarks.

In this paper, we explore the space of running gaits for the bipedal robot Cassie. Our first contribution is to present an approach for optimizing gait efficiency across a spectrum of speeds with the aim of enabling extremely high-speed running on hardware. This raises the question of how the resulting gaits compare to human running mechanics, which are known to be highly efficient in comparison to quadrupeds. Our second contribution is to conduct this comparison based on established human biomechanical studies. We find that despite morphological differences between Cassie and humans, key properties of the gaits are highly similar across a wide range of speeds. Finally, our third contribution is to integrate the optimized running gaits into a full controller that satisfies the rules of the real-world task of the 100m dash, including starting and stopping from a standing position. We demonstrate this controller on hardware to establish the Guinness World Record for Fastest 100m by a Bipedal Robot.

Recent work on sim-to-real learning for bipedal locomotion has demonstrated new levels of robustness and agility over a variety of terrains. However, that work, and most prior bipedal locomotion work, have not considered locomotion under a variety of external loads that can significantly influence the overall system dynamics. In many applications, robots will need to maintain robust locomotion under a wide range of potential dynamic loads, such as pulling a cart or carrying a large container of sloshing liquid, ideally without requiring additional load-sensing capabilities. In this work, we explore the capabilities of reinforcement learning (RL) and sim-to-real transfer for bipedal locomotion under dynamic loads using only proprioceptive feedback. We show that prior RL policies trained for unloaded locomotion fail for some loads and that simply training in the context of loads is enough to result in successful and improved policies. We also compare training specialized policies for each load versus a single policy for all considered loads and analyze how the resulting gaits change to accommodate different loads. Finally, we demonstrate sim-to-real transfer, which is successful but shows a wider sim-to-real gap than prior unloaded work, which points to interesting future research.

Recently, work on reinforcement learning (RL) for bipedal robots has successfully learned controllers for a variety of dynamic gaits with robust sim-to-real demonstrations. In order to maintain balance, the learned controllers have full freedom of where to place the feet, resulting in highly robust gaits. In the real world however, the environment will often impose constraints on the feasible footstep locations, typically identified by perception systems. Unfortunately, most demonstrated RL controllers on bipedal robots do not allow for specifying and responding to such constraints. This missing control interface greatly limits the real-world application of current RL controllers. In this paper, we aim to maintain the robust and dynamic nature of learned gaits while also respecting footstep constraints imposed externally. We develop an RL formulation for training dynamic gait controllers that can respond to specified touchdown locations. We then successfully demonstrate simulation and sim-to-real performance on the bipedal robot Cassie. In addition, we use supervised learning to induce a transition model for accurately predicting the next touchdown locations that the controller can achieve given the robot's proprioceptive observations. This model paves the way for integrating the learned controller into a full-order robot locomotion planner that robustly satisfies both balance and environmental constraints.

In this work, we propose a learning approach for 3D dynamic bipedal walking when footsteps are constrained to stepping stones. While recent work has shown progress on this problem, real-world demonstrations have been limited to relatively simple open-loop, perception-free scenarios. Our main contribution is a more advanced learning approach that enables real-world demonstrations, using the Cassie robot, of closed-loop dynamic walking over moderately difficult stepping-stone patterns. Our approach first uses reinforcement learning (RL) in simulation to train a controller that maps footstep commands onto joint actions without any reference motion information. We then learn a model of that controller's capabilities, which enables prediction of feasible footsteps given the robot's current dynamic state. The resulting controller and model are then integrated with a real-time overhead camera system for detecting stepping stone locations. For evaluation, we develop a benchmark set of stepping stone patterns, which are used to test performance in both simulation and the real world. Overall, we demonstrate that sim-to-real learning is extremely promising for enabling dynamic locomotion over stepping stones. We also identify challenges remaining that motivate important future research directions.

Recent work has demonstrated the success of reinforcement learning (RL) for training bipedal locomotion policies for real robots. This prior work, however, has focused on learning joint-coordination controllers based on an objective of following joint trajectories produced by already available controllers. As such, it is difficult to train these approaches to achieve higher-level goals of legged locomotion, such as simply specifying the desired end-effector foot movement or ground reaction forces. In this work, we propose an approach for integrating knowledge of the robot system into RL to allow for learning at the level of task space actions in terms of feet setpoints. In particular, we integrate learning a task space policy with a model-based inverse dynamics controller, which translates task space actions into joint-level controls. With this natural action space for learning locomotion, the approach is more sample efficient and produces desired task space dynamics compared to learning purely joint space actions. We demonstrate the approach in simulation and also show that the learned policies are able to transfer to the real bipedal robot Cassie. This result encourages further research towards incorporating bipedal control techniques into the structure of the learning process to enable dynamic behaviors.

We propose a method for optimizing control policies for assistive lower-limb devices. The method frames parameter selection as a dueling bandits problem in which a user indicates his or her qualitative preferences between pairs of parameter sets chosen from a library. We generate the library through an offline optimization procedure that seeks to reproduce the varied gaits of healthy human subjects. By separating the parameter selection process into online and offline portions, the method can handle high-dimensional parameter spaces and produces policies that can generalize to different gait scenarios such as speed variation. We evaluate the method on five subjects walking on a powered knee and ankle prosthesis governed by a neuromuscular control policy that has 43 parameters. We find the five subjects preferred four different parameter sets from the library and that the resulting optima resemble intact subject gait data. This result suggests the offline portion of the optimization method indeed produces control parameters that can adapt to different gaits. Moreover, we find that for three out of the four parameter sets we tested, the procedure also generates parameters that improve the ability of the prosthesis to adapt to increasing gait speed by increasing ankle net work production. The results encourage further research and exploration in clinical settings toward advanced prosthesis controls that employ online learning.