What’s next for offshore energy?


Dr Tomasz Luczynski and Dr Jonatan Scharff Willners are Research Associates for Robotics and Autonomous Systems at Heriot-Watt University.


On December 7th, they will present the latest subsea robotics innovations at the webinar ‘Robotics and AI: What’s Next for Offshore Energy?’ that is jointly organised by ORCA Hub, ORE Catapult and Net Zero Technology Centre.

With the increased development of new offshore infrastructures, such as wind turbines, offshore farming and oil&gas infrastructure, the need for frequent inspection and maintenance of those assets is greater than ever. This is conventionally performed by remotely operated vehicles (ROVs). Offshore ROV operations are expensive, as they require to cover the cost of the support vessel, skilled crew and often have to be repeated or are completed only partially due to unfavourable weather conditions. Any measures allowing for faster completion of the inspection and automated assessment of the collected data will provide great savings.

To address this, we have been developing a system to perform much of the prior manual work autonomously. We aim to replace the support vessel with an autonomous surface vehicle (AUV) which can launch and recover the ROV autonomously. The ROV is endowed with software to make it capable of performing inspections of the environment in cooperation with an operator or completely autonomously.

Our ideas have been developed and validated within the ORCA Hub - a multimillion-pound research project that addresses the challenges of remote monitoring, inspection, and intervention of offshore assets using robots and autonomous systems. It is also backed by many years of experience in marine robotics and now reached the stage when the developed algorithms are ready to be commercialised.

By moving the personnel from being on-site to a remote location, much of the human risk is mitigated and the size of the support vessels can be reduced - leading to a smaller cost and environmental impact. Likewise, without humans at sea, the work can be performed in harsher conditions which means less downtime or aborted missions. As the ROV can perform missions autonomously, a single person can operate multiple platforms simultaneously. If the operator needs to intervene, they can easily take control to the degree they like, this could be to inject a new target to inspect or completely control the vehicle. For the majority of operations, the end goal is to analyse the data to determine if any additional actions are needed. Our system can generate data in a form that makes these decisions easier, such as correct scale detailed 3D models. Analysing a 3D model is an easier process that reduces the risk of missing something important, compared to scanning through hours of video.

We aim to deliver platform agnostic autonomy and perception packages that can be deployed on commercially produced robots. In addition, we offer our custom payload that supplies the robot with the necessary hardware required to enable our software packages. Our offering can be split into several distinct modules:

1. Image enhancement – software providing real-time image enhancement, that offers clear advantages over normal video:

  • Risk reduction – clearer image provides better situational awareness and reduces the risk of crashing into the structure, tether entanglement or getting stuck in tight spaces.
  • Flexibility – inspections can be performed in a wider range of conditions, even when the visibility wouldn't allow for collecting good data with the traditional vision system.
  • Cost reduction – Inspections can be carried out faster, more reliably and efficiently.
Image enhancement

2. SLAM – simultaneous localisation and mapping module, providing real-time information about the motion of the vehicle and current position in the map. Additionally, the SLAM module creates an added value by:

  • More reliable localisation – reduces the risk of the pilot getting lost and shortens the time required to find the desired place.
  • Improved understanding of the asset being inspected.
  • Improved safety by enabling collision-free navigation.
  • Enables autonomy to react to and engage with surroundings.

3. 3D modelling – thanks to the SLAM module, a 3D model can be created allowing for:

  • Easy understanding of the data collected – rather than watching long videos, the asset can be inspected as a 3D model. This is not only faster and more natural way of assessing the collected data, but also creates a spatial awareness of the inspected asset.
  • Enables automated reporting – 3D model can be used to automatically generate the mission report and compare the collected data against models saved during previous missions. It’s not only faster and more in-depth analysis, but also greatly reduces the risk of human error while assessing the state of the structure from the video.
3D modelling

4. Autonomy package – a set of advanced control and planning algorithms that allow for much easier vehicle control and autonomous or semi-autonomous inspection. The key benefits of using this package are:

  • Cost reduction – less qualified/trained pilots are required. Furthermore, the pilot does not have to be physically on board, given an LTE/satellite connection to the vessel. This also enables full unmanned operations if the surface vessel can deploy and recover the ROV.
  • Better quality control – the algorithm can ensure full coverage of the inspected region.
  • Semi-automated operations – a single operator can simultaneously supervise multiple vehicles.
  • Reduced carbon emission through more efficient operations and reduction of the vessel time.
  • The frequency can be increased without increasing the cost, thus improving the safety and understanding of the asset. Maintenance scheduling can also be optimised thanks to the data collected from more frequent inspections.
  • Health and safety improvements through reduced offshore personnel.
Autonomy package

5. Smart cameras and underwater embedded processing units – all our libraries can be offered with custom-designed hardware which is easily integrated on an ROV and can be expanded. Adding new functionalities will only require software updates or a small hardware expansion unit that can be deployed as plug-and-play. A fully-featured hardware platform with integrated software will be offered as Autonomous Underwater Inspection Payload (AUIP)