Aidan Curtis

I'm a fifth year PhD student at MIT EECS advised by Leslie Kaelbling, Tomás Lozano-Pérez, and Joshua Tenenbaum. I am grateful to be supported by the NSF Graduate Research Fellowship and the Harold E. Edgerton Memorial Fellowship. My recent internship experiences include IBM, MERL, and Boston Dynamics.

My work spans the fields of robotics, artificial intelligence, machine learning, reasoning, and planning. The focus of my PhD has been on long-horizon manipulation problems where reasoning under uncertainty and partial observability is crucial. I'm passionate about bringing advanced reasoning and decision-making capabilities to real-world robots.

Email  /  Google Scholar  /  Twitter  /  Github

Domain randomization in reinforcement learning is an established technique for increasing the robustness of control policies learned in simulation. In this paper, we present a more flexible representation for domain randomization using normalizing flows, and show how the learned flows can be used as artifacts for multi-step planning.

We prompt the LLM to output code for a function with open parameters, which, together with environmental constraints, can be viewed as a Continuous Constraint Satisfaction Problem (CCSP). This CCSP can be solved through sampling or optimization to find a skill sequence and continuous parameter settings that achieve the goal while avoiding constraint violations.

We present Bayes3D, an uncertainty-aware perception system for structured 3D scenes, that reports accurate posterior uncertainty over 3D object shape, pose, and scene composition in the presence of clutter and occlusion. Bayes3D delivers these capabilities via a novel hierarchical Bayesian model for 3D scenes and a GPU-accelerated coarse-to-fine sequential Monte Carlo algorithm.

Generalizable long-horizon robotic assembly requires reasoning at multiple levels of abstraction. We propose a hierarchical modular approach, named ARCH which combines imitation learning and reinforcement learning for long-horizon high-precision assembly in contact-rich settings.

When using sampling-based motion planners such as PRMs, it is difficult to determine how many samples are required for the PRM to find a solution consistently. We attempt to address this problem by proving an upper bound on the number of samples that are sufficient, with high probability, for a radius PRM to find a feasible solution, drawing on prior work in deterministic sampling and sample complexity theory.

Integrated task and motion planning (TAMP) has proven to be a valuable approach to generalizable long-horizon robotic manipulation and navigation problems. We propose a strategy for TAMP with Uncertainty and Risk Awareness (TAMPURA) that is capable of efficiently solving long-horizon planning problems with initial-state and action outcome uncertainty, including problems that require information gathering and avoiding undesirable and irreversible outcomes.

In this paper, we propose STRUG, an online POMDP solver capable of handling domains that require long-horizon planning with significant task-relevant and task-irrelevant uncertainty. We demonstrate our solution on several temporally extended versions of toy POMDP problems as well as robotic manipulation of articulated objects using neural perception.

In this paper, we examine the problem of visibility aware robot navigation among movable obstacles (VANAMO). A variant of the well-known NAMO robotic planning problem, VANAMO puts additional visibility constraints on robot motion and object movability.

In this work, we study generalized policy search-based methods with a focus on the score function used to guide the search over policies. The main idea is that a candidate policy should be used to guide planning on training problems as a mechanism for evaluating that candidate.

This workshop paper describes our award-winning solution to the Sapien Maniskill Manipulation Challenge. We present a framework for building generalizable manipulation controller policies that map from raw input point clouds and segmentation masks to joint velocities.

We present a strategy for designing and building very general robot manipulation systems involving the integration of a general-purpose task-and-motion planner with engineered and learned perception modules that estimate properties and affordances of unknown objects.

Generalized planning accelerates classical planning by finding an algorithm-like policy that solves multiple instances of a task. Here we apply generalized planning to hybrid discrete-continuous task and motion planning.

Humans are expert explorers. In this work, we try to understand the computational cognitive mechanisms that support this efficiency can advance the study of the human mind and enable more efficient exploration algorithms.

Real-world planning problems often involve hundreds or even thousands of objects, straining the limits of modern planners. In this work, we address this challenge by learning to predict a small set of objects that, taken together, would be sufficient for finding a plan.

Reading words aloud is a foundational aspect of the acquisition of literacy. We used direct intracranial recordings in a large cohort to create a holistic yet fine-grained map of word processing, enabling us to derive the spatiotemporal neural codes of multiple word attributes critical to reading

A core problem of long-range planning is finding an efficient way to search through the tree of possible action sequences. Here, we propose the Curious Sample Planner (CSP), which fuses elements of TAMP and DRL by combining a curiosity-guided sampling strategy with imitation learning to accelerate planning.

We introduce ThreeDWorld (TDW), a platform for interactive multi-modal physical simulation.

We take a software-inspired viewpoint of clinical trial designs to enable expressibility of complex trials, composability with diverse devices and services while maximally maintaining simplicity for a clinical research user.

Top-down visual object recognition processes driven by the human orbitofrontal cortex (OFC) have been proposed to facilitate rapid processing of images in higher-level visual regions. We found that trials lacking the saccade artifact on scalp EEG also lacked low gamma band PLV increase in iEEG. This work illustrates the importance of eliminating confounding saccadic artifacts.

We use Deep Reinforcement Learning to train AI agents which are able to combat wildfires. This page demonstrates the videos of our learned policies. Please see our paper for more details.

Predicting short term video dynamics has many useful applications in self-driving cars, weather nowcasting, and model-based reinforcement learning. In this project we provide an in-depth analysis of the available models for video prediction and their strengths and weaknesses in predicting natural sequences of images.

The project assesses the efficacy of finetuning off different neural architectures trained on ImageNet for Actor-Critic reinforcement learning in 2D and 3D environments. We find a combination of semantic and spatial information results in the best few-shot performance.

In this project we design and build a wireless intracranial neural recording system that uses sparse coding compression to efficiently transmit neural data.

website source code