CausalWorld: A Robotic Manipulation Benchmark for Causal Structure and Transfer Learning

By Ossama Ahmed, Frederik Trauble, Anirudh Goyal, Alexander Neitz, M. Wuthrich, Yoshua Bengio, Bernhard Schölkopf, Stefan Bauer. 2020

149 citations recorded by the source corpus.

Despite recent successes of reinforcement learning (RL), it remains a challenge for agents to transfer learned skills to related environments. To facilitate research addressing this problem, we propose CausalWorld, a benchmark for causal structure and transfer learning in a robotic manipulation environment. The environment is a simulation of an open-source robotic platform, hence offering the possibility of sim-to-real transfer. Tasks consist of constructing 3D shapes from a given set of blocks - inspired by how children learn to build complex structures. The key strength of CausalWorld is that it provides a combinatorial family of such tasks with common causal structure and underlying factors (including, e.g., robot and object masses, colors, sizes). The user (or the agent) may intervene on all causal variables, which allows for fine-grained control over how similar different tasks (or task distributions) are. One can thus easily define training and evaluation distributions of a desired difficulty level, targeting a specific form of generalization (e.g., only changes in appearance or object mass). Further, this common parametrization facilitates defining curricula by interpolating between an initial and a target task. While users may define their own task distributions, we present eight meaningful distributions as concrete benchmarks, ranging from simple to very challenging, all of which require long-horizon planning as well as precise low-level motor control. Finally, we provide baseline results for a subset of these tasks on distinct training curricula and corresponding evaluation protocols, verifying the feasibility of the tasks in this benchmark.

Subfields: Artificial Intelligence, Control and Systems Engineering, Causal Inference & Discovery, Robotics & Control

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Control and Systems Engineering

In Causal Inference & Discovery

In Robotics & Control

Citations in the corpus

Toward Causal Representation Learning (2021, background): “An emerging line of work devoted to the intersection of RL and causality includes [1], [13], [22], [38], [51], [165], [276].”
Towards Causal Representation Learning (2021, background): “An emerging line of work devoted to the intersection of RL and causality includes [13, 21, 164, 37, 50, 275, 1].”
Continual World: A Robotic Benchmark For Continual Reinforcement Learning (2021, background): “[3] Ossama Ahmed, Frederik Träuble, Anirudh Goyal, Alexander Neitz, Manuel Wuthrich, Yoshua Bengio, Bernhard Schölkopf, and Stefan Bauer.”
Towards Continual Reinforcement Learning: A Review and Perspectives (2020, background): “Ahmed et al. (2020) have recently proposed CausalWorld, which is a promising benchmark fulfilling many of these criteria in application to robotic manipulation tasks.”
Inductive biases for deep learning of higher-level cognition (2020, background): “This complicates the task of evaluating and comparing learning algorithms but is necessary and we believe, feasible, e.g. see Yu et al. (2017); Packer et al. (2018); Chevalier-Boisvert et al. (2018); Dulac-Arnold et al. (2020); Ahmed et al. (2020).”
Instruction-driven history-aware policies for robotic manipulations (2022, background): “The emergence of robotic simulators, such as Gym [52], manipulaTHOR [53], dm control [54], Sapien [55], CausalWorld [56], and RL-Bench [19], also greatly accelerated the development of manipulation methods.”
D’ya Like DAGs? A Survey on Structure Learning and Causal Discovery (2021, methodology): “[139]) user specified adds edges with probability p = 2e d2−d Linear, GP Add, GP Mix, Sigmoid Add and Sigmoid Mix mixed graph data CausalWorld [2] comprehensive robotics dataset MPI3D [71] visual disentanglement dataset Pendulum-light-shadow [257] image data Phase coupled oscillator [7] physical rel”
VIMA: General Robot Manipulation with Multimodal Prompts (2022, background): “CausalWorld (Ahmed et al., 2021) is a benchmark for causal structure and transfer learning in manipulation, requiring long-horizon planning and precise low-level motor control.”
LIBERO: Benchmarking Knowledge Transfer for Lifelong Robot Learning (2023, unknown): “Robot Learning Benchmarks A variety of robot learning benchmarks have been proposed to address challenges in meta learning (MetaWorld [73]), causality learning (CausalWorld [1]), multi-task learning [27, 35], policy generalization to unseen objects [47, 24], and compositional learning [44].”
Systematic Evaluation of Causal Discovery in Visual Model Based Reinforcement Learning (2021, background): “CausalWorld Ahmed et al. [2020] consists of a suite of robotics tasks, it allows the user to control some parameters of the underlying causal graph, such as the number of variables and the attributes of variables (such as size and weight).”
How to Reuse and Compose Knowledge for a Lifetime of Tasks: A Survey on Continual Learning and Functional Composition (2022, background): “Another related work procedurally created robotics tasks by varying dynamical parameters to study causality in RL (Ahmed et al., 2021), considering a single robot arm dealing with continuous variations in the physical properties of objects.”