Pyrobot: An effort by Facebook

Reading Time: 3 minutes |

September 20, 2022

|

AI

WhatNext

Pyrobot: An effort by Facebook

With growing interest in the area of machine learning and automation, there has been a wave of development in the area of robotics. Most people in the tech community and elsewhere have come across the common parlance regarding “robots stealing our jobs”. However, the currently available category of robots suffers a major drawback. These robots lack in the area of precision and control which is claimed to be possible only through highly efficient software control.

The field of robotics is not new. In the last two decades, attempts from robot researchers have resulted in a series of improved robots. With the emergence of machine learning, the ways in which a robot learns about its surroundings is changing rapidly. The technological developments in this area are clustered across the world and ground-breaking research is taking place in silos. The development of these technologies demands a deep knowledge base with the expertise to come up with new applications. Tech majors have been trying to grab a pie of this development by developing their ecosystem of these technologies.

Facebook has shown a keen interest in the area of robotics. Recently, it open sourced its simulator AI Habitat. The company claims that its simulator can be used to train AI agents (hardware and software) that are used in consumer spaces and require training to understand the behavior of its fellow space occupants.

Recently, it developed Pyrobot, in association with the Carnegie Mellon University. It is a ROS based interface that provides users with a range of APIs, for the development and control of various features in a robot. It helps users to run deep learning models trained by Pytorch (Facebook’s machine learning platform). The idea is to have a central knowledge base for AI developers, researchers and students to enable them to get robots working without specialized knowledge of different aspects of navigation, grasping and torque control, etc. This enables them to experiment with new concepts and ideas. It can also lead to the development of high-end AI and robotics applications without the need to worry about the base layer.

Similar efforts have been made by Amazon with the Robomaker in the past. The Robomaker is an interface based on ROS. Microsoft showed its interest by developing a custom solution development kit. Further, it integrated ROS with its Windows 10 platform to enable widespread adoption.

The democratization of robotics can be understood from the approach of not establishing own base layers (by technology majors), but by using the already omnipresent ROS. Such interface systems (such as pyrobot) developed by cash-rich big players can mark an important milestone in the area of technology research. This is mainly because these steps reduce the barriers to entry in this field which is otherwise limited (to an extent) to big research universities and companies. It becomes further easy to use and deploy with low-cost systems such as the LoCobot. Researchers can deploy several experimental robots, swarms, etc. that collect data over time. Such initiatives can unearth ground-breaking results that can have significant implications for current and future industrial and consumer users.

Facebook mentions that it has its focus on the education industry. This can ensure early familiarity with the AI and robotics concepts among students leading to a career in this field. Pyrobot enables users to write simple codes in python that are then translated into complex instructions that a typical ROS program requires. Without such systems, it used to be difficult and time consuming for programmers to develop and test new concepts.

The field of robotics is catching up with consumer and industrial applications. These intelligent devices are increasingly being used across applications. With automation levels reaching their peaks, it won’t be an exaggeration to say that robots will become more pervasive in industrial/consumer spaces. The amount of data captured by these robots will be massive. Facebook, with its focus on user data, aims to grab a pie of the vast data-set that is set to be generated by new age robots.

However, it is very early to comment on the utility of pyrobot and similar ROS interfaces. Currently, the interface works on two robots- Sawyer and LoCoBot, a low-cost mobile robot with a 5-DOF manipulator. It can be a challenge for pyrobot to be universally used across all systems. To develop a hardware agnostic solution is one of the challenges, Facebook is trying to solve with its platform. In the robotics community, this is not the first attempt at the unification of technological developments. The problem here is that similar attempts in the past have seen interest from roboticists and students, but this interest has lasted for a small time.

Researchers move to a parallel platform with better/more suitable features. With increasing security threats with autonomous systems and the recent data leak news that tarnished its reputation, it will be interesting to see how Facebook tackles security issues associated with its new system.

Overall, with new features, pyrobot is a strong attempt from the social network giant at streamlining robotic and AI research while enabling scholars with testing of deep learning models without domain expertise. In the coming days, it is expected to be used across different robot systems, further scaling its adoption in the area of research.

To deep dive and stay continuously updated about the most recent global innovations in Artificial Intelligence and learn more about applications in your industry, test drive WhatNext now!

How useful was this post?

Click on a star to rate it!

Average rating 0 / 5. Vote count: 0

No votes so far! Be the first to rate this post.

Leave a Comment

Your email address will not be published. Required fields are marked *

Related Insights

Food Supply Chain - WhatNext

Food Supply Chain and Internet of Things

Driver Monitoring using AI -WhatNext

Driver Monitoring using Artificial Intelligence

Quantum Computing - WhatNext

Quantum Computing in Car Manufacturing

Sustainable Agriculture - WhatNext

Sustainable Agriculture using Synthetic Biology

Potential of Living Medicines - WhatNext

Potential of Living Medicines