Autonomous robot navigation

During the summer of 2016, I worked on application development, computer vision, and robotics for an autonomous robot navigation startup. My work primarily centered on improving the autonomous and semi-autonomous navigation capabilities of mobile robots, particularly the "follow-me" function. This functionality is critical for enabling robots to assist users by carrying heavy items or learning specific navigation paths through user-guided training, which can later be executed autonomously.

For mobile robots, safe and efficient navigation is essential. Key capabilities include localization—determining the robot's position and orientation within a given frame of reference—and path planning, which involves creating a route to a goal location while avoiding collisions and unsafe conditions such as extreme temperatures, uneven surfaces, or radiation exposure. The follow-me function, while valuable, posed significant challenges in its existing implementations. Traditional systems often relied on detecting changes in distance or orientation between the robot and the user. However, these systems were overly sensitive, leading to unnecessary and unnatural motions, inefficient battery usage, and abrupt movements that could pose safety risks to nearby individuals.

To address these limitations, I contributed to the development of a more natural and efficient follow-me solution. This new approach enabled mobile robots to shadow user movements smoothly, reducing unnecessary adjustments and abrupt directional changes. These improvements enhanced user safety, optimized energy efficiency, and significantly improved the overall user experience. By mitigating the drawbacks of traditional systems, this work demonstrated the potential for robotics to integrate seamlessly into everyday environments. For further details, please refer to the related patent application, which outlines the technical innovations and impact of this solution.

See patent application for more details.

Using C++ and OpenCV, I developed a decision-making module enabling a robot to autonomously track a moving person. In addition to designing and implementing this core functionality, I collaborated with engineers and researchers to improve and rigorously test the code. My contributions extended beyond technical development to include:

Technology Evangelism: Delivered demonstrations and presentations to stakeholders, investors, and vendors, effectively communicating the value and potential of the system.

Product Development and Launch: Partnered with the design team to develop an Android application, ensuring seamless integration of the technology and enhancing user experience.

Impact and Significance

This project represents a step forward in making autonomous robotics more practical, user-friendly, and scalable. It bridges the gap between theoretical research and real-world implementation, contributing to the broader goal of integrating robotics into everyday life to improve efficiency, safety, and quality of life.