Interacting with human agents in complex scenarios presents a significant challenge for robotic navigation, particularly in environments that necessitate both collision avoidance and collaborative interaction, such as indoor spaces. Unlike static or predictably moving obstacles, human behavior is inherently complex and unpredictable, stemming from dynamic interactions with other agents. Existing simulation tools frequently fail to adequately model such reactive and collaborative behaviors, impeding the development and evaluation of robust social navigation strategies. This paper introduces a novel framework utilizing distributed potential games to simulate human-like interactions in highly interactive scenarios. Within this framework, each agent imagines a virtual cooperative game with others based on its estimation. We demonstrate this formulation can facilitate the generation of diverse and realistic interaction patterns in a configurable manner across various scenarios. Additionally, we have developed a gym-like environment leveraging our interactive agent model to facilitate the learning and evaluation of interactive navigation algorithms.

Reinforcement Learning (RL) has the potential to enable robots to learn from their own actions in the real world. Unfortunately, RL can be prohibitively expensive, in terms of on-robot runtime, due to inefficient exploration when learning from a sparse reward signal. Designing dense reward functions is labour-intensive and requires domain expertise. In our work, we propose GCR (Goal-Contrastive Rewards), a dense reward function learning method that can be trained on passive video demonstrations. By using videos without actions, our method is easier to scale, as we can use arbitrary videos. GCR combines two loss functions, an implicit value loss function that models how the reward increases when traversing a successful trajectory, and a goal-contrastive loss that discriminates between successful and failed trajectories. We perform experiments in simulated manipulation environments across RoboMimic and MimicGen tasks, as well as in the real world using a Franka arm and a Spot quadruped. We find that GCR leads to a more-sample efficient RL, enabling model-free RL to solve about twice as many tasks as our baseline reward learning methods. We also demonstrate positive cross-embodiment transfer from videos of people and of other robots performing a task. Appendix: \url{https://tinyurl.com/gcr-appendix-2}.

In reinforcement learning for legged robot locomotion, crafting effective reward strategies is crucial. Pre-defined gait patterns and complex reward systems are widely used to stabilize policy training. Drawing from the natural locomotion behaviors of humans and animals, which adapt their gaits to minimize energy consumption, we propose a simplified, energy-centric reward strategy to foster the development of energy-efficient locomotion across various speeds in quadruped robots. By implementing an adaptive energy reward function and adjusting the weights based on velocity, we demonstrate that our approach enables ANYmal-C and Unitree Go1 robots to autonomously select appropriate gaits, such as four-beat walking at lower speeds and trotting at higher speeds, resulting in improved energy efficiency and stable velocity tracking compared to previous methods using complex reward designs and prior gait knowledge. The effectiveness of our policy is validated through simulations in the IsaacGym simulation environment and on real robots, demonstrating its potential to facilitate stable and adaptive locomotion.

Automating the assembly of objects from their parts is a complex problem with innumerable applications in manufacturing, maintenance, and recycling. Unlike existing research, which is limited to target segmentation, pose regression, or using fixed target blueprints, our work presents a holistic multi-level framework for part assembly planning consisting of part assembly sequence inference, part motion planning, and robot contact optimization. We present the Part Assembly Sequence Transformer (PAST) -- a sequence-to-sequence neural network -- to infer assembly sequences recursively from a target blueprint. We then use a motion planner and optimization to generate part movements and contacts. To train PAST, we introduce D4PAS: a large-scale Dataset for Part Assembly Sequences (D4PAS) consisting of physically valid sequences for industrial objects. Experimental results show that our approach generalizes better than prior methods while needing significantly less computational time for inference.

Designing robotic agents to perform open vocabulary tasks has been the long-standing goal in robotics and AI. Recently, Large Language Models (LLMs) have achieved impressive results in creating robotic agents for performing open vocabulary tasks. However, planning for these tasks in the presence of uncertainties is challenging as it requires \enquote{chain-of-thought} reasoning, aggregating information from the environment, updating state estimates, and generating actions based on the updated state estimates. In this paper, we present an interactive planning technique for partially observable tasks using LLMs. In the proposed method, an LLM is used to collect missing information from the environment using a robot and infer the state of the underlying problem from collected observations while guiding the robot to perform the required actions. We also use a fine-tuned Llama 2 model via self-instruct and compare its performance against a pre-trained LLM like GPT-4. Results are demonstrated on several tasks in simulation as well as real-world environments. A video describing our work along with some results could be found here.

This paper introduces a full system modeling strategy for a syringe pump and soft pneumatic actuators(SPAs). The soft actuator is conceptualized as a beam structure, utilizing a second-order bending model. The equation of natural frequency is derived from Euler's bending theory, while the damping ratio is estimated by fitting step responses of soft pneumatic actuators. Evaluation of model uncertainty underscores the robustness of our modeling methodology. To validate our approach, we deploy it across four prototypes varying in dimensional parameters. Furthermore, a syringe pump is designed to drive the actuator, and a pressure model is proposed to construct a full system model. By employing this full system model, the Linear-Quadratic Regulator (LQR) controller is implemented to control the soft actuator, achieving high-speed responses and high accuracy in both step response and square wave function response tests. Both the modeling method and the LQR controller are thoroughly evaluated through experiments. Lastly, a gripper, consisting of two actuators with a feedback controller, demonstrates stable grasping of delicate objects with a significantly higher success rate.

Contact-rich manipulation is common in a wide range of robotic applications, including assembly [1, 2, 3, 4, 5, 6, 7, 8], object pivoting [9, 10, 11], grasping[12, 13, 14], and pushing [15, 16]. To accomplish these tasks, robots need to learn both the manipulation trajectory and the force control parameters. The manipulation trajectory guides the robot toward completing the task while physically engaging with the environment, whereas the force control parameters regulate the contact force. Incorrect control parameters can lead to oscillations and excessive contact forces that may damage the robot or the environment. Past works have tackled the contact-rich skill-learning problem in different ways. First, the majority of previous works [2, 10, 6, 7, 3, 17, 9, 18, 19] focus on learning the manipulation trajectories and rely on human experts to manually tune force control parameters. While this simplification has demonstrated remarkable performance in many applications, letting human labor tune control parameters is still inconvenient. Furthermore, the tuned parameter for one task may not generalize well to other task settings with different kinematic or dynamic properties. For example, assembly tasks with different clearances will require different control parameters.

Trajectory planning is crucial for the safe driving of autonomous vehicles in highway traffic flow. Currently, some advanced trajectory planning methods utilize spatio-temporal voxels to construct feasible regions and then convert trajectory planning into optimization problem solving based on the feasible regions. However, these feasible region construction methods cannot adapt to the changes in dynamic environments, making them difficult to apply in complex traffic flow. In this paper, we propose a trajectory planning method based on adaptive spatio-temporal voxels which improves the construction of feasible regions and trajectory optimization while maintaining the quadratic programming form. The method can adjust feasible regions and trajectory planning according to real-time traffic flow and environmental changes, realizing vehicles to drive safely in complex traffic flow. The proposed method has been tested in both open-loop and closed-loop environments, and the test results show that our method outperforms the current planning methods.

Local planning for a differential wheeled robot is designed to generate kinodynamic feasible actions that guide the robot to a goal position along the navigation path while avoiding obstacles. Reactive, predictive, and learning-based methods are widely used in local planning. However, few of them can fit static and crowd environments while satisfying kinodynamic constraints simultaneously. To solve this problem, we propose a novel local planning method. The method applies a long-term dynamic window approach to generate an initial trajectory and then optimizes it with graph optimization. The method can plan actions under the robot's kinodynamic constraints in real time while allowing the generated actions to be safer and more jitterless. Experimental results show that the proposed method adapts well to crowd and static environments and outperforms most SOTA approaches.

Humans inherently possess generalizable visual representations that empower them to efficiently explore and interact with the environments in manipulation tasks. We advocate that such a representation automatically arises from simultaneously learning about multiple simple perceptual skills that are critical for everyday scenarios (e.g., hand detection, state estimate, etc.) and is better suited for learning robot manipulation policies compared to current state-of-the-art visual representations purely based on self-supervised objectives. We formalize this idea through the lens of human-oriented multi-task fine-tuning on top of pre-trained visual encoders, where each task is a perceptual skill tied to human-environment interactions. We introduce Task Fusion Decoder as a plug-and-play embedding translator that utilizes the underlying relationships among these perceptual skills to guide the representation learning towards encoding meaningful structure for what's important for all perceptual skills, ultimately empowering learning of downstream robotic manipulation tasks. Extensive experiments across a range of robotic tasks and embodiments, in both simulations and real-world environments, show that our Task Fusion Decoder consistently improves the representation of three state-of-the-art visual encoders including R3M, MVP, and EgoVLP, for downstream manipulation policy-learning. Project page: https://sites.google.com/view/human-oriented-robot-learning