One of Facebook AI Research’s long term goals is to work together with the AI research community to advance AI and build intelligent agents. These agents need to model and understand the physical world they live in. We believe robots can play a critical role in exploring, modeling, and learning about the physical world. In recent years, we have seen significant advancements in both robotic hardware and accompanying algorithms. However, due to expensive hardware and the lack of research ecosystem and platforms, it has been extremely challenging (a) for non-robotics researchers to demonstrate application of ideas on robotics problems, (b) to benchmark and evaluate performance of robotics algorithms, and (c) to quickly identify, disseminate, and improve upon ideas that work well. In order to democratize robotics research, Facebook AI Research, in collaboration with Carnegie Mellon University, has introduced an open-source research platform: PyRobot.
PyRobot is a light-weight, high-level interface on top of ROS that provides a consistent set of hardware independent mid-level APIs to control different robots. PyRobot abstracts away details about low-level controllers and inter-process communication, and allows non-robotics researchers (ML, CV researchers) to focus on building high-level AI applications. PyRobot aims to provide a research ecosystem with convenient access to robotics datasets, algorithm implementations, and models that can be used to quickly create a state-of-the-art baseline.
To promote the first use of the library and faster adaptation, we are pleased to accept proposals focused on using PyRobot with the recently introduced low-cost hardware LoCoBot. The proposals should focus on how the applicants plan to use PyRobot and demonstrate new tasks, help the robotics research community by implementing benchmark algorithms. Applicants from the academic community are invited to submit a maximum one-page proposal outlining their intended use of the LoCoBot and PyRobot library, and estimated timeline.