Machine Learning for Robotics

Book description

Machine learning is the present and future of robotics, whether it's for self-driving vehicles, consumer robotics, or industrial manufacturing. Driven by breakthroughs in research and compute infrastructure, widespread deployment of huge neural nets, and end-to-end robotics tasks, the science and practice of robotics is in the midst of disruption from data-driven approaches such as deep learning and foundation models.

If you're a software or machine learning engineer looking to get into robotics, or a robotics engineer planning to deploy machine learning in your projects, this is your book.

You'll learn how to apply deep learning methods to robotics and approach core robotics technologies—perception, reasoning, and prediction—from a deep learning perspective. This guide explores state-of-the-art deep learning algorithms relevant to each core technology and shows you how to use them for real-world robotics. Relevant code samples demonstrate how to apply these algorithms.

You'll learn how to:

  • Formulate robotics as a data-driven AI problem
  • Recognize the technology behind designing and deploying modern robotics: sensing, perception, training, and control
  • Apply state-of-the-art techniques in AI to robotics systems
  • Understand factors driving decision-making in technical design for several robotics applications
  • Design practical robotic systems for real-world applications: self-driving, prosthetics, and industrial automation

Publisher resources

View/Submit Errata

Product information

  • Title: Machine Learning for Robotics
  • Author(s): Alishba Imran, Keerthana Gopalakrishnan
  • Release date: June 2024
  • Publisher(s): O'Reilly Media, Inc.
  • ISBN: 9781098134198