Perception Engineer - NPI Robotics

Role Overview

Dyson is on the lookout for candidates with direct experience in applied robotics and a robust portfolio showcasing their robotics projects. We value individuals who demonstrate initiative, creativity, and a rapid learning ability, complemented by a diverse skill set that can enhance our company’s mission.

• Design, develop, and assess innovative sensing solutions for Dyson's next generation of robotic platforms, tackling real-world robotic challenges.
• Create prototypes in both simulation and hardware environments to conduct comprehensive quantitative evaluations to determine their suitability and robustness.
• Identify, analyze, and address the complex edge cases encountered in real-world domestic settings.
• Collaborate within a multidisciplinary team to engineer systems that consider the physical, computational, and financial constraints of the target platform.
• Write software using C++ and Python, deploying and testing it on prototype or target hardware.
• Engineer cutting-edge machine learning and image processing algorithms tailored for specific hardware.
• Stay informed about emerging technologies in the robotics domain and remain updated on industry best practices and trends.

Candidate Profile

• Hold an advanced degree (or possess equivalent industry experience) in robotics or a related field, emphasizing the application of advanced concepts to practical challenges in robotic sensing and perception.
• Have a minimum of 2 years of experience working with a variety of proprioceptive and exteroceptive sensing technologies, understanding their features, limitations, and integration methods.
• Possess practical experience with various Time-of-Flight and RGB camera sensors.
• Demonstrate experience in designing, developing, deploying, and testing both hardware and software prototypes for proof-of-concept initiatives.
• Proficient in the implementation of leading machine learning operations (MLOps) and computer vision algorithms, including but not limited to: object identification, object tracking, point cloud registration, and image segmentation.
• Have experience with robotics simulation platforms such as Gazebo or PyBullet.
• Exhibit strong programming capabilities in C++, Python, and ROS (Robot Operating System), along with familiarity with relevant open-source libraries and development tools.
• Have experience in developing and deploying software solutions within Linux environments, employing modern software development best practices.
• Able to articulate complex concepts in a clear and concise manner.