Challenge

Lack of robotic-platform-agnostic autonomous inspection systems.

Solution

The development of an automated pipeline for robotic inspection and data management that enables changes to the robotic platform, sensor suite, mission parameters and data management platform without fundamental re-engineering. Autonomous inspection for hazardous environments has become a proven application of robotics. However, rapid scaling of this technology is blocked by the need to embark on the scratch development of a new robotics system to meet the specific requirements of each use case; this slows adoptions and limits application to only the highest value use cases.

The on-going AutoInspect project from Createc uses the Rooster, a sensor pack developed by Oxford Robotics Institute (ORI) for mapping, localisation and navigation purposes. It implements an algorithm known as Pronto, which is a versatile and modular EKF state estimator used for both proprioceptive (inertial, kinematics) and exteroceptive (LIDAR, camera) sensor fusion. This enables the system to generate an accurate 3D map of the robot’s surroundings, while fusing the robot’s odometry with IMU and camera data to generate a prediction of the robot’s location in the map.