Project Description

Introduction

As more companies begin to invest in technology to automate simple tasks, it isn’t a question of when automation will affect a given company, but how it will be implemented. Should a device be created to fulfill the tasks/services of a human operator within the current infrastructure? Or should the entire infrastructure be replaced with updated, computer controlled devices? In the naval industry, the latter approach is embraced by research company Leidos in their autonomous unmanned surface vehicle the Sea Hunter. This ship has revolutionized the way we think about sea navigation, and begged the question: Can the systems that power and maintain the Sea Hunter be translated to a large-scale cargo ship?

Leidos is at the forefront of this research, and they currently believe it will be more cost-effective to develop an autonomous robot that will help perform maintenance on the various electro-mechanical devices found on a cargo ship, rather than designing and building new cargo ships with the Sea Hunter’s infrastructure. Our team was tasked with developing this robot, which is able to manipulate all valves and breakers on a mock-up testbed. With RoboDutchman, we took a minimalist approach to cut down cost and complexity, investing more time into developing a robust planning algorithm. By leveraging the Robot Operating System, we’ve integrated various sensors and controllers that allow for accurate localization in any environment, fast object-detection/classification, and optimized task execution. This platform allows for seamless integration of other sensors/software that can improve the future performance of the robot.

Specification Sheet