Fire Service College, October 2018

Over 50 research staff from across each of ORCA’s partner institutions participated in the second ORCA capability demonstrations which focused on terrestrial and aerial activities.

A team was invited from the University of Oxford to represent our sister Hub, RAIN (Robotics and AI in Nuclear Research) was also invited to take part. We were delighted to have strong attendance from industry, in addition to our EPSRC representative and the Independent Steering Committee Chair.

Out at the College’s oil rig site, human-robot interaction lead, Prof. Helen Hastie from Heriot-Watt University gave attendees an indoor demonstration of a scenario showing MIRIAM providing emergency response to an unmanned aerial vehicle (UAV) which was inspecting a fire on an offshore oil rig in a simulated environment.

Photos from the demsonstrations at the Fire Service College, October 2018

At the rig site, the University of Oxford team showcased their Dynamic Robot Systems (DRS) research on legged mobility for inspection using the ANYmal quadruped and a remote operation software stack which allows the operator to supervise the progress of the mission and plan the robot’s next actions. Starting at a distance from the oil rig, ANYmal demonstrated its ability to localise its position relative to a prior 3D model of the environment, after which it was tasked with reaching a target area inside the rig. This involved navigation across uneven tarmac and the traversing of a 25cm barrier for which the team inverted its leg configuration. The team then used a speedy trotting gait to reach an industrial staircase which it climbed using a walking gait before entering the rig to explore.

The University of Edinburgh team led by Dr. Michael Mistry gave their ANYmal demonstration showing semi-autonomous sensing and intervention with mobile robots by manoeuvring ANYmal around both seen and unseen obstacles on top of the rig where it also had to contend with different challenging surfaces.

Dr Adam Stokes and his group from the University of Edinburgh discussed the further developments of the Limpet sensor. Future demonstrations will include communication from the Limpets back to a central station and show their monitoring capabilities either through showing real-time data or adaptable edge processing.

Photos from the rig site demonstrations at Fire Service College.

Dr. Chang Liu and Dr. Pouyan Khalili from Imperial College London gave demonstrations of autonomous sensor deployment using a large drone to carry the sensor from ground level to land on a specified point on top of the rig.

Using manned aviation practice as inspiration, the University of Liverpool team are working towards demonstrating the use of distributed simulation that can be used to generate evidence to help certify both offshore assets and the assets that gather the data themselves. Led by Dr. Mike Jump, the Liverpool team demonstrated the progress that has so far been made including an initial distributed simulation example.

Colleagues from the RAIN Hub at the University of Oxford demonstrated a repeatable autonomous inspection routine using visual teach and repeat on a Husky wheeled robot on top of the rig. This was followed up with a demonstration of long-term inspection planning under hazardous conditions using a Jackal wheeled robot autonomously navigating around the base of the oil rig.

Prof. Helen Hastie:

"We very much value our industry support and working closely with our industry partners because they allow us to understand what the real problems are that operators and users face, and create real solutions for these real problems and move the technology further."