Autonomous Industrial Inspection in a Contextualized Environment

Asset-heavy industries such as Oil and Gas, shipping and utilities have long sought after increased automation of inspection, maintenance and repair (IMR). Such tasks are generally tedious, time-consuming and repetitive. Moreover, IMR-tasks often expose industry personnel to health, safety and environmental risks due to the hazardous nature of industrial sites. The recent commercial availability of flexible industry-grade robotic platforms such as the Boston Dynamics Spot and EX?Robotics?ExR-1 opens new opportunities for autonomous IMR. The use of such platforms along with so-called digital twins of the industrial sites to be inspected is poised to catalyse automation of IMR-tasks. Digital twins hold up to date digital representations of the industrial sites in question. A digital twin typically includes data such as 3D models, images and live sensor data. Robots can be a reliable source of data to the digital twin, and thereby increase its value. Meanwhile the digital twin can provide situational awareness to the robots, and thereby improving their navigation and behaviour when conducting IMR-tasks. In this project we strive to further the state of autonomous inspection, i.e. inspection done by robots without human supervision, by exploiting synergies between mature robotic platforms?and digital twins of industrial sites. The overarching goal is to develop theory on autonomous IMR informed by the situational awareness provided by the digital?twin, and?subsequently test the developed methods in real-world settings.

-

Asset-heavy industries such as Oil and Gas, shipping and utilities have long sought after increased automation of inspection, maintenance and repair (IMR). Such tasks are generally tedious, time-consuming and repetitive. IMR-tasks often expose industry personnel to health, safety and environmental risks due to the hazardous nature of industrial sites. Additionally, automated IMR is a prerequisite for the goal of truly unmanned operation of industrial sites. The use of autonomous mobile robots has shown great promise in this regard. The industry is still in its infancy with respect to autonomous IMR. The advent of mature robotic platforms and digital twins of industrial sites is poised to catalyse the development of autonomous IRM. Flexible robotic platforms such as the Boston Dynamics Spot, EX Robotics ExR-1 and ANYbotics ANYmal opens new opportunities for autonomous IMR. Industry actors can disregard complicated and expensive hard ware aspects of autonomous robotics and focus on providing high level IMR-functionality on top of the capabilities innate to the platforms. Paired with digital twins, the autonomous robots have a context to operate within. The digital twin provides the robots with increased situational awareness. Additionally the robots increase the value of the digital twin by keeping it up to date and providing data that can be used to increase operational efficiency. In this project we strive to further the state of autonomous IMR by exploiting synergies between autonomous robotics and digital twins of industrial sites. The primary objective is to achieve robust autonomous inspection of industrial sites by building on state-of-the-art methods from autonomous exploration and mapping and improving them by exploiting the interconnection with the digital twin.