We investigate policy transfer using image-to-semantics translation to mitigate learning difficulties in vision-based robotics control agents. This problem assumes two environments: a simulator environment with semantics, that is, low-dimensional and essential information, as the state space, and a real-world environment with images as the state space. By learning mapping from images to semantics, we can transfer a policy, pre-trained in the simulator, to the real world, thereby eliminating real-world on-policy agent interactions to learn, which are costly and risky. In addition, using image-to-semantics mapping is advantageous in terms of the computational efficiency to train the policy and the interpretability of the obtained policy over other types of sim-to-real transfer strategies. To tackle the main difficulty in learning image-to-semantics mapping, namely the human annotation cost for producing a training dataset, we propose two techniques: pair augmentation with the transition function in the simulator environment and active learning. We observed a reduction in the annotation cost without a decline in the performance of the transfer, and the proposed approach outperformed the existing approach without annotation.

Despite the popularity of policy gradient methods, they are known to suffer from large variance and high sample complexity. To mitigate this, we introduce SoftTreeMax -- a generalization of softmax that takes planning into account. In SoftTreeMax, we extend the traditional logits with the multi-step discounted cumulative reward, topped with the logits of future states. We consider two variants of SoftTreeMax, one for cumulative reward and one for exponentiated reward. For both, we analyze the gradient variance and reveal for the first time the role of a tree expansion policy in mitigating this variance. We prove that the resulting variance decays exponentially with the planning horizon as a function of the expansion policy. Specifically, we show that the closer the resulting state transitions are to uniform, the faster the decay. In a practical implementation, we utilize a parallelized GPU-based simulator for fast and efficient tree search. Our differentiable tree-based policy leverages all gradients at the tree leaves in each environment step instead of the traditional single-sample-based gradient. We then show in simulation how the variance of the gradient is reduced by three orders of magnitude, leading to better sample complexity compared to the standard policy gradient. On Atari, SoftTreeMax demonstrates up to 5x better performance in a faster run time compared to distributed PPO. Lastly, we demonstrate that high reward correlates with lower variance.

We propose policy gradient algorithms which learn risk-sensitive policies in a reinforcement learning (RL) framework. Our proposed algorithms maximize the distortion risk measure (DRM) of the cumulative reward in an episodic Markov decision process in on-policy as well as off-policy RL settings. We derive a variant of the policy gradient theorem that caters to the DRM objective, and use this theorem in conjunction with a likelihood ratio-based gradient estimation scheme. We derive non-asymptotic bounds that establish the convergence of our proposed algorithms to an approximate stationary point of the DRM objective.

As human-robot collaboration increases in the workforce, it becomes essential for human-robot teams to coordinate efficiently and intuitively. Traditional approaches for human-robot scheduling either utilize exact methods that are intractable for large-scale problems and struggle to account for stochastic, time varying human task performance, or application-specific heuristics that require expert domain knowledge to develop. We propose a deep learning-based framework, called HybridNet, combining a heterogeneous graph-based encoder with a recurrent schedule propagator for scheduling stochastic human-robot teams under upper- and lower-bound temporal constraints. The HybridNet's encoder leverages Heterogeneous Graph Attention Networks to model the initial environment and team dynamics while accounting for the constraints. By formulating task scheduling as a sequential decision-making process, the HybridNet's recurrent neural schedule propagator leverages Long Short-Term Memory (LSTM) models to propagate forward consequences of actions to carry out fast schedule generation, removing the need to interact with the environment between every task-agent pair selection. The resulting scheduling policy network provides a computationally lightweight yet highly expressive model that is end-to-end trainable via Reinforcement Learning algorithms. We develop a virtual task scheduling environment for mixed human-robot teams in a multi-round setting, capable of modeling the stochastic learning behaviors of human workers. Experimental results showed that HybridNet outperformed other human-robot scheduling solutions across problem sizes for both deterministic and stochastic human performance, with faster runtime compared to pure-GNN-based schedulers.

We propose a Model-Based Reinforcement Learning (MBRL) algorithm named VF-MC-PILCO, specifically designed for application to mechanical systems where velocities cannot be directly measured. This circumstance, if not adequately considered, can compromise the success of MBRL approaches. To cope with this problem, we define a velocity-free state formulation which consists of the collection of past positions and inputs. Then, VF-MC-PILCO uses Gaussian Process Regression to model the dynamics of the velocity-free state and optimizes the control policy through a particle-based policy gradient approach. We compare VF-MC-PILCO with our previous MBRL algorithm, MC-PILCO4PMS, which handles the lack of direct velocity measurements by modeling the presence of velocity estimators. Results on both simulated (cart-pole and UR5 robot) and real mechanical systems (Furuta pendulum and a ball-and-plate rig) show that the two algorithms achieve similar results. Conveniently, VF-MC-PILCO does not require the design and implementation of state estimators, which can be a challenging and time-consuming activity to be performed by an expert user.

The complexity of designing reward functions has been a major obstacle to the wide application of deep reinforcement learning (RL) techniques. Describing an agent's desired behaviors and properties can be difficult, even for experts. A new paradigm called reinforcement learning from human preferences (or preference-based RL) has emerged as a promising solution, in which reward functions are learned from human preference labels among behavior trajectories. However, existing methods for preference-based RL are limited by the need for accurate oracle preference labels. This paper addresses this limitation by developing a method for crowd-sourcing preference labels and learning from diverse human preferences. The key idea is to stabilize reward learning through regularization and correction in a latent space. To ensure temporal consistency, a strong constraint is imposed on the reward model that forces its latent space to be close to the prior distribution. Additionally, a confidence-based reward model ensembling method is designed to generate more stable and reliable predictions. The proposed method is tested on a variety of tasks in DMcontrol and Meta-world and has shown consistent and significant improvements over existing preference-based RL algorithms when learning from diverse feedback, paving the way for real-world applications of RL methods.

Offline reinforcement learning aims to train a policy on a pre-recorded and fixed dataset without any additional environment interactions. There are two major challenges in this setting: (1) extrapolation error caused by approximating the value of state-action pairs not well-covered by the training data and (2) distributional shift between behavior and inference policies. One way to tackle these problems is to induce conservatism - i.e., keeping the learned policies closer to the behavioral ones. To achieve this, we build upon recent works on learning policies in latent action spaces and use a special form of Normalizing Flows for constructing a generative model, which we use as a conservative action encoder. This Normalizing Flows action encoder is pre-trained in a supervised manner on the offline dataset, and then an additional policy model - controller in the latent space - is trained via reinforcement learning. This approach avoids querying actions outside of the training dataset and therefore does not require additional regularization for out-of-dataset actions. We evaluate our method on various locomotion and navigation tasks, demonstrating that our approach outperforms recently proposed algorithms with generative action models on a large portion of datasets.

Slip is a very common phenomena present in wheeled mobile robotic systems. It has undesirable consequences such as wasting energy and impeding system stability. To tackle the challenge of mobile robot trajectory tracking under slippery conditions, we propose a hierarchical framework that learns and adapts gains of the tracking controllers simultaneously online. Concretely, a reinforcement learning (RL) module is used to auto-tune parameters in a lateral predictive controller and a longitudinal speed PID controller. Experiments show the necessity of simultaneous gain tuning, and have demonstrated that our online framework outperforms the best baseline controller using fixed gains. By utilizing online gain adaptation, our framework achieves robust tracking performance by rejecting slip and reducing tracking errors when the mobile robot travels through various terrains.

Reinforcement Learning (RL; Sutton and Barto, 2018; Mannor et al., 2022) studies online decision making problems in which an agent learns through experience within a dynamic environment, with the goal to minimize a loss function associated with the agent-environment interaction. Modern applications of RL such as robotics(Schulman et al., 2015; Lillicrap et al., 2015; Akkaya et al., 2019), game playing (Mnih et al., 2013; Silver et al., 2018) and autonomous driving (Kiran et al., 2021), almost invariably consist of large scale environments where function approximation techniques are necessary to allow the agent to generalize across different states. Furthermore, some form of agent robustness is usually required to cope with environment irregularities that cannot be faithfully represented by stochasticity assumptions (see e.g., Dulac-Arnold et al., 2021). Theoretical foundations for RL with function approximation (e.g., Jiang et al., 2017; Yang and Wang, 2019; Jin et al., 2020b; Agarwal et al., 2020) have been steadily coming into fruition.

Many instances of similar or almost-identical industrial machines or tools are often deployed at once, or in quick succession. For instance, a particular model of air compressor may be installed at hundreds of customers. Because these tools perform distinct but highly similar tasks, it is interesting to be able to quickly produce a high-quality controller for machine $N+1$ given the controllers already produced for machines $1..N$. This is even more important when the controllers are learned through Reinforcement Learning, as training takes time, energy and other resources. In this paper, we apply Policy Intersection, a Policy Shaping method, to help a Reinforcement Learning agent learn to solve a new variant of a compressors control problem faster, by transferring knowledge from several previously learned controllers. We show that our approach outperforms loading an old controller, and significantly improves performance in the long run.