Use of Machine Learning for Helicopter Ship Operational Research

The Flight Science and Technology (FS&T) research group at the University of Liverpool are international leaders in the use of flight simulation to support aircraft/ship and helicopter/ship clearance operations. This PhD offers an excellent opportunity for a highly motivated student to work with the UK MoD, and the international research community, on a project that could have a significant impact on the way real-world operations are conducted in the future. The launch and recovery of helicopters to/from a ship are challenging operations due to the combination of a confined ship landing deck, irregular ship motion, sea spray and unsteady airflow over and around the ship’s landing deck and superstructure. For each new combination of helicopter and ship, evidence is required to determine the acceptable pilot workload prior to safe operating limits being reached. This evidence is usually derived from flight trials and a Ship Helicopter Operating Limit (SHOL) diagram is generated from these trials indicating the magnitude and direction of winds that can be safely flown in during helicopter launch and recovery operations.

The UK MoD has invested a significant amount of time, effort and cost in rotorcraft clearance activities, generating a wealth of flight and simulation test data. To advance the state of the art in SHOL development, a structured examination of the contributing factors to pilot workload in maritime environments, both at-sea and in-flight simulation, is required. This research has the potential to significantly reduce the time and cost of future clearances.

The objective of this study is to develop a “smart” approach to support future aircraft clearance activities through a thorough examination of existing flight test and simulation SHOL data. Real-world and simulation datasets are expensive assets that have not been fully examined to understand, analyse and predict aircraft operational limits. This project will develop a new methodology to interrogate existing data and identify sensitivities in the modelling and simulation (M&S) environment to a range of operational conditions. It will generate new workload metrics and best practices for future aircraft clearance activities to improve the efficiency of SHOL work and could potentially improve operational capability of existing and future ship/helicopter combinations. Fulfilling this objective will represent a significant change in best practices and will have potential applications in the clearance processes for future optionally piloted and unmanned platforms.

The following activities are envisaged for the project:

1. Review current Helicopter/Ship Operational practices: A placement at Dstl will provide the student with valuable insights into the methods for planning, conducting, analysis and reporting flight/simulator SHOL trials. If possible, the student would seek involvement in future at-sea trials to gain first-hand experience of this process. This phase of the project will lead to a definition of the best practice test methodology for gathering real-world and simulation data to enable comparative analyses to be conducted.
2. The “smart” approach is to use Machine Learning (ML) to identify patterns in existing datasets with minimal human intervention. There are a range of ML techniques available for examining textual and numerical data and the student will review these to identify the most appropriate method for this study.
3. FS&T has generated a large dataset from simulated/flight SHOL trials which will be utilised in this ML assessment stage; similar real-word data will be sought from MoD. Cataloguing of datasets will be required to identify gaps in existing data representing different operational conditions to identify future simulation trial requirements.
4. The subjective assessment provided by the pilot in SHOL testing, at-sea or in simulation, is key in the determination of the SHOL. The research will examine the concept of benchmark testing or calibrating pilots to determine pilot sensitivity to different environmental/simulator conditions. An examination of pilot workload, linking subjective opinion and the new workload objective metrics, will be conducted to perform a sensitivity analysis of different simulator elements e.g. motion cueing, to define new simulator trials.
5. The research will aim to develop an offline tool for SHOL predictions using pilot modelling techniques, enabling a wide range of test conditions to be examined and will help optimise future at-sea and simulation trials.

This research will draw upon existing NATO AVT activities that FS&T are involved with to maintain the UK’s position at the forefront of M&S SHOL providing the successful applicant with a unique opportunity to work at an international level. The position is open to UK nationals primarily but applications from EU nationals are also invited.

To apply for this opportunity, please visit: https://www.liverpool.ac.uk/study/postgraduate-research/how-to-apply/

For any enquiries please contact Professor White on [Email Address Removed]