Collaborative Robot Autonomy and Localisation (CORAL)

Project

Heterogeneous multi-agent robotic inspection missions

Benefits

Reduced risk to humans and improved repeatability, accuracy and availability of monitoring.

Current Status

The project is ongoing and currently focusing on the underwater vehicle.

Delivery Partners

The University of Manchester

Contact Details

Mel Willis

technical.innovation@sellafieldsites.com

Wet and dry storage facilities must be routinely monitored to ensure that corrosion, leaks, or degradation of facilities is identified rapidly. Unfortunately, this is not an easy task for human operators due to the presence of active materials. The CORAL project aims to automate these inspection tasks that would otherwise be performed by humans, minimising the risk of humans receiving a radiation dose.

Two robot teams that can perform repeatable autonomous inspections of wet and dry facilities without the

need for additional infrastructure are being developed. The two robot teams comprise:

1. An aquatic surface vehicle and an underwater vehicle.

2. A ground vehicle that can deploy an Unmanned Aerial Vehicle (UAV).

A requirement for mobile robot autonomy is a positioning system. The complex arrangement of most nuclear facilities, and the fact that coverage is limited by the placement of the sensors, means that installing positioning system infrastructure is often infeasible.

The CORAL project solves this issue by using a multi robot team. One of the robots is used as a movable base to position from i.e. the aquatic

surface vehicle and ground vehicle. These robots self-localise using LiDAR and Simultaneous Localisation and Mapping technology. The base robots also compute the real time position of the partner robots (i.e., the UAV and underwater vehicle) by tracking active markers placed directly on the partner robot.

Using more than one robot allows for fast and accurate robot positioning in environments where Global Positioning System (GPS) is not available, without the need for any additional infrastructure. The accurate tracking of the robot’s position will also enable them to revisit exact positions to repeat measurements. This collaborative approach also enables consistent, long-term monitoring so that potential problems can be identified early and rectified in the safest and most cost effective way.