Abstract
The development of capable subsea vehicles to perform autonomous inspection missions is nearing the point where deployment and operational use is becoming a reality. The challenge for the widespread deployment is initially the acceptance of autonomous vehicles to perform missions and to be confident in the performance of the vehicle.
The suitability of the near shore testing areas (playgrounds) to de-risk the performance of the vehicle, is starting to be understood, and it is broadly expected that the gap between emulating an offshore environment in a playground, and performing autonomous missions on an operational site, will require support. Today, this is typically addressed by the presence of a crewed surface vessel providing selective status information to the operating crew on board the vessel due to the limitation of data transfer through water.
Use of widespread subsea based communication networks could provide a meshed network subsea, improving the available data, but this is challenging for a brownfield application, especially in a Life Extension Context due to anticipated limitations in the supporting Subsea Production System electrical infrastructure.
In support of an ambition to develop geo-referencing of a resident autonomous underwater vehicle, operating without the aid of a surface vessel, an onboard Digital Field Map, DFM, capability is relevant to overcome those limitations. The DFM provides the opportunity to encode Mission Intelligence into the meta data of the DFM. The use of Mission Intelligence potentially reduces the complexity of Artificial Intelligence and offers an opportunity to establish clear acceptance criteria. The definition of acceptance criteria presents a clear demonstration of the validation of the actions of the vehicle rather than a subjective assessment of the performance of the vehicle under Artificial Intelligence control.
In addition, the DFM also presents an opportunity to be used for Mission Planning, providing a visualization of the intended mission profile and to define and implement mitigations identified at a Risk Assessment directly into the vehicle without the risk of loss of detail when implementing mitigations in code, as well as simplifying the validation of the implementation of the mitigating steps.
