CoBRA: A Composable Benchmark for Robotics Applications

TL;DR

CoBRA introduces a comprehensive benchmark suite for evaluating automated robot configuration and planning across various tasks, environments, and robot types, facilitating standardized comparison and development in robotics optimization.

Contribution

It provides a unified, modular benchmark platform for robotics applications, especially aiding the development of automated optimization methods for complex robot configurations.

Findings

Benchmark suite covers diverse tasks and environments

Accessible platform for sharing and comparing solutions

Supports modular robots with many parameters

Abstract

Selecting an optimal robot, its base pose, and trajectory for a given task is currently mainly done by human expertise or trial and error. To evaluate automatic approaches to this combined optimization problem, we introduce a benchmark suite encompassing a unified format for robots, environments, and task descriptions. Our benchmark suite is especially useful for modular robots, where the multitude of robots that can be assembled creates a host of additional parameters to optimize. We include tasks such as machine tending and welding in synthetic environments and 3D scans of real-world machine shops. All benchmarks are accessible through https://cobra.cps.cit.tum.de, a platform to conveniently share, reference, and compare tasks, robot models, and solutions.