This paper focuses on the problem of detecting and reacting to changes in the distribution of a sensorimotor controller's observables. The key idea is the design of switching policies that can take conformal quantiles as input, which we define as conformal policy learning, that allows robots to detect distribution shifts with formal statistical guarantees. We show how to design such policies by using conformal quantiles to switch between base policies with different characteristics, e.g. safety or speed, or directly augmenting a policy observation with a quantile and training it with reinforcement learning. Theoretically, we show that such policies achieve the formal convergence guarantees in finite time. In addition, we thoroughly evaluate their advantages and limitations on two compelling use cases: simulated autonomous driving and active perception with a physical quadruped. Empirical results demonstrate that our approach outperforms five baselines. It is also the simplest of the baseline strategies besides one ablation. Being easy to use, flexible, and with formal guarantees, our work demonstrates how conformal prediction can be an effective tool for sensorimotor learning under uncertainty.

Moving objects to find a fully-occluded target object, known as mechanical search, is a challenging problem in robotics. As objects are often organized semantically, we conjecture that semantic information about object relationships can facilitate mechanical search and reduce search time. Large pretrained vision and language models (VLMs and LLMs) have shown promise in generalizing to uncommon objects and previously unseen real-world environments. In this work, we propose a novel framework called Semantic Mechanical Search (SMS). SMS conducts scene understanding and generates a semantic occupancy distribution explicitly using LLMs. Compared to methods that rely on visual similarities offered by CLIP embeddings, SMS leverages the deep reasoning capabilities of LLMs. Unlike prior work that uses VLMs and LLMs as end-to-end planners, which may not integrate well with specialized geometric planners, SMS can serve as a plug-in semantic module for downstream manipulation or navigation policies. For mechanical search in closed-world settings such as shelves, we compare with a geometric-based planner and show that SMS improves mechanical search performance by 24% across the pharmacy, kitchen, and office domains in simulation and 47.1% in physical experiments. For open-world real environments, SMS can produce better semantic distributions compared to CLIP-based methods, with the potential to be integrated with downstream navigation policies to improve object navigation tasks. Code, data, videos, and the appendix are available: https://sites.google.com/view/semantic-mechanical-search

Grasping objects by a specific part is often crucial for safety and for executing downstream tasks. Yet, learning-based grasp planners lack this behavior unless they are trained on specific object part data, making it a significant challenge to scale object diversity. Instead, we propose LERF-TOGO, Language Embedded Radiance Fields for Task-Oriented Grasping of Objects, which uses vision-language models zero-shot to output a grasp distribution over an object given a natural language query. To accomplish this, we first reconstruct a LERF of the scene, which distills CLIP embeddings into a multi-scale 3D language field queryable with text. However, LERF has no sense of objectness, meaning its relevancy outputs often return incomplete activations over an object which are insufficient for subsequent part queries. LERF-TOGO mitigates this lack of spatial grouping by extracting a 3D object mask via DINO features and then conditionally querying LERF on this mask to obtain a semantic distribution over the object with which to rank grasps from an off-the-shelf grasp planner. We evaluate LERF-TOGO's ability to grasp task-oriented object parts on 31 different physical objects, and find it selects grasps on the correct part in 81% of all trials and grasps successfully in 69%. See the project website at: lerftogo.github.io

We consider a decluttering problem where multiple rigid convex polygonal objects rest in randomly placed positions and orientations on a planar surface and must be efficiently transported to a packing box using both single and multi-object grasps. Prior work considered frictionless multi-object grasping. In this paper, we introduce friction to increase the number of potential grasps for a given group of objects, and thus increase picks per hour. We train a neural network using real examples to plan robust multi-object grasps. In physical experiments, we find a 13.7% increase in success rate, a 1.6x increase in picks per hour, and a 6.3x decrease in grasp planning time compared to prior work on multi-object grasping. Compared to single-object grasping, we find a 3.1x increase in picks per hour.

The AlphaGarden is an automated testbed for indoor polyculture farming which combines a first-order plant simulator, a gantry robot, a seed planting algorithm, plant phenotyping and tracking algorithms, irrigation sensors and algorithms, and custom pruning tools and algorithms. In this paper, we systematically compare the performance of the AlphaGarden to professional horticulturalists on the staff of the UC Berkeley Oxford Tract Greenhouse. The humans and the machine tend side-by-side polyculture gardens with the same seed arrangement. We compare performance in terms of canopy coverage, plant diversity, and water consumption. Results from two 60-day cycles suggest that the automated AlphaGarden performs comparably to professional horticulturalists in terms of coverage and diversity, and reduces water consumption by as much as 44%. Code, videos, and datasets are available at https://sites.google.com/berkeley.edu/systematiccomparison.

Amazon, Nimble, Plus One, Waymo, and Zoox use remote human supervision of robot fleets in applications ranging from self-driving taxis to automated warehouse fulfillment [1, 2, 3, 4, 5]. These robots intermittently cede control during task execution to remote human supervisors for corrective interventions. The interventions take place either during learning, when they are used to improve the robot policy, or during execution, when the policy is no longer updated but robots can still request human assistance when needed to improve reliability. In the continual learning setting, these occur simultaneously: the robot policy has been deployed but continues to be updated indefinitely with additional intervention data. Furthermore, any individual robot can share its intervention data with the rest of the fleet. As opposed to robot swarms that must coordinate with each other to achieve a common objective, a robot fleet is a set of independent robots simultaneously executing the same control policy in parallel environments. We refer to the setting of a robot fleet learning via interactive requests for human supervision (see Figure 1) as Interactive Fleet Learning (IFL). Of central importance in IFL is the supervisor allocation problem: how should limited human supervision be allocated to robots in a manner that maximizes the throughput of the fleet?

Deep ensembles have been shown to extend the positive effect seen in typical ensemble learning to neural networks and to reinforcement learning (RL). However, there is still much to be done to improve the efficiency of such ensemble models. In this work, we present Diverse Ensembles for Fast Transfer in RL (DEFT), a new ensemble-based method for reinforcement learning in highly multimodal environments and improved transfer to unseen environments. The algorithm is broken down into two main phases: training of ensemble members, and synthesis (or fine-tuning) of the ensemble members into a policy that works in a new environment. The first phase of the algorithm involves training regular policy gradient or actor-critic agents in parallel but adding a term to the loss that encourages these policies to differ from each other. This causes the individual unimodal agents to explore the space of optimal policies and capture more of the multimodality of the environment than a single actor could. The second phase of DEFT involves synthesizing the component policies into a new policy that works well in a modified environment in one of two ways. To evaluate the performance of DEFT, we start with a base version of the Proximal Policy Optimization (PPO) algorithm and extend it with the modifications for DEFT. Our results show that the pretraining phase is effective in producing diverse policies in multimodal environments. DEFT often converges to a high reward significantly faster than alternatives, such as random initialization without DEFT and fine-tuning of ensemble members. While there is certainly more work to be done to analyze DEFT theoretically and extend it to be even more robust, we believe it provides a strong framework for capturing multimodality in environments while still using RL methods with simple policy representations.

Polyculture farming has environmental advantages but requires substantially more pruning than monoculture farming. We present novel hardware and algorithms for automated pruning. Using an overhead camera to collect data from a physical scale garden testbed, the autonomous system utilizes a learned Plant Phenotyping convolutional neural network and a Bounding Disk Tracking algorithm to evaluate the individual plant distribution and estimate the state of the garden each day. From this garden state, AlphaGardenSim selects plants to autonomously prune. A trained neural network detects and targets specific prune points on the plant. Two custom-designed pruning tools, compatible with a FarmBot gantry system, are experimentally evaluated and execute autonomous cuts through controlled algorithms. We present results for four 60-day garden cycles. Results suggest the system can autonomously achieve 0.94 normalized plant diversity with pruning shears while maintaining an average canopy coverage of 0.84 by the end of the cycles. For code, videos, and datasets, see https://sites.google.com/berkeley.edu/pruningpolyculture.

Abstract-- This paper presents AlphaGarden: an autonomous polyculture garden that prunes and irrigates living plants in a 1.5m 3.0m physical testbed. AlphaGarden uses an overhead camera and sensors to track the plant distribution and soil moisture. We model individual plant growth and interplant dynamics to train a policy that chooses actions to maximize leaf coverage and diversity. For autonomous pruning, AlphaGarden uses two custom-designed pruning tools and a trained neural network to detect prune points. Results suggest AlphaGarden can autonomously achieve 0.96 normalized diversity with pruning shears while maintaining an average canopy coverage of 0.86 during the peak of the cycle. Industrial agriculture is based on monoculture, where a single crop type is cultivated, requiring substantial use of, pesticides, and water [1], [2].

The difficulty in specifying rewards for many real-world problems has led to an increased focus on learning rewards from human feedback, such as demonstrations. However, there are often many different reward functions that explain the human feedback, leaving agents with uncertainty over what the true reward function is. While most policy optimization approaches handle this uncertainty by optimizing for expected performance, many applications demand risk-averse behavior. We derive a novel policy gradient-style robust optimization approach, PG-BROIL, that optimizes a soft-robust objective that balances expected performance and risk. To the best of our knowledge, PG-BROIL is the first policy optimization algorithm robust to a distribution of reward hypotheses which can scale to continuous MDPs. Results suggest that PG-BROIL can produce a family of behaviors ranging from risk-neutral to risk-averse and outperforms state-of-the-art imitation learning algorithms when learning from ambiguous demonstrations by hedging against uncertainty, rather than seeking to uniquely identify the demonstrator's reward function.