Statement of Aim
This project aims to develop the simulation model of a helicopter test pilot: to implement the behaviour of a human helicopter test pilot into desktop simulations of helicopters operating at sea.
Background
Doing simulations from a desktop computer requires a much simpler setup than real world trials, whether in flight simulators or at sea. A desktop pilot model can provide support to real-world trials as a preview assessment of their safety to fly can help prioritise the testing conditions to consider during real world trials. Implementing a digital test pilot model into the desktop simulation stage can help mitigate the risks, improve efficiency and lower the costs involved with real world trials.
Companies commercialising helicopters have the responsibility to determine and demonstrate the capabilities of the vehicles they develop. The safety limits of a helicopter operating to a ship depend on the aircraft’s response, the pilot’s capabilities to counteract them, and the operating conditions. They are traditionally defined during at-sea flight trials on a ‘test and declare’ basis. Test pilots perform repeated launch and recovery missions from and to a ship, and then assess the safety of the manoeuvre. The operational clearance envelope is established from the flight conditions that are deemed safe, and to provide the largest possible operational clearance envelope, flight trials need to consider as many environmental conditions as possible. However, due to the unpredictable nature of at-sea test conditions, it can result in an incomplete and restricted operational envelope, achieved at very high financial cost and at significant risk to the crew.
Problem
The UK MoD has been funding research and development to investigate the use of modelling and simulation in the context of helicopters operating at sea for the past 15 years, with the aim of reducing the number of trials at sea required for initial clearance to fly a helicopter to a ship.
However, most MoD work to date has been based on implementing piloted simulations: complex simulation models focusing on capturing aircraft behaviour, ship motion and environment are implemented into flight simulators. The tasks simulated are then piloted by human pilots. While this approach has the potential to reduce the number of trials at sea required, running piloted simulations is still costly and time consuming.
A better way to approach the problem would be to consider offline simulations from a desktop computer along with piloted simulations and trials at sea. The aim is to use offline simulation to reduce the number of piloted simulations and subsequently further reduce the number of trials at sea required. However, current industry practices implementing desktop simulations of helicopters operating at sea lack a pilot model that adequately captures the behaviour of a human pilot.
Industrial partnership
Nova Systems is a leader in providing engineering services and technology solutions for flight tests. Large flight datasets are acquired to provide evidence for aircraft development, certification and acceptance. Modern aircrafts have a wealth of data, whose potential is mostly untapped, for both upfront evidence generation and through-life management. The ability to exploit that data could reap large cost benefits and extract further capability in optimised design, maintenance and usage
The development of a test pilot model using these experimental data with digital algorithms and machine learning techniques is thus of high interest to Nova Systems. However, access to data scientists is challenging in the aerospace industry. Indeed, it must compete with often more attractive sectors such as technology, business and finance.
Hence, the partnership between Nova Systems and the Distributed Algorithms CDT to develop a novel process to inform real-world helicopter ship testing with the implementation of a digital helicopter pilot model.
The CDT provides highly trained data scientists addressing the needs of industry for individuals able to develop and deliver high-performance products and services through the use of data-driven optimisation and high-performance computing.
For Nova Systems, the project addresses the challenges related to the extensive costs and time requirements of traditional live flight testing for ship-helicopter operations with the development of an advanced alternative to make operations at sea safer and more cost-effective.
In exchange, the project benefits from Nova Systems’ extensive experience in flight test, certification and acceptance of aircraft and flight simulators.
The project is ongoing with the development of a first pilot simulation model and the collection of experimental data from a range of pilots. Data is acquired with piloted trials using the unique flight simulator facilities of the University of Liverpool. It is then implemented in the pilot model developed.
"Nova Systems are excited to be involved in this opportunity to push the boundaries of digital engineering and distributed algorithms. We believe the skillsets and pedigree of the partnership between Nova Systems and the University of Liverpool are uniquely matched to offer future students to be uniquely employable in the growing domain of Digital Engineering in the Aerospace sector."
Student
As an aerospace engineer with a special focus on flight dynamics, I was uniquely suited for this project. I have industry experience with implementing simulation models in aviation as a former Flight Control Engineer for the UK-developed novelty aircraft PHASA-35. I was responsible for the aircraft’s simulation model and algorithms development for the autopilot so that the aircraft could gain clearance to fly unmanned in the stratosphere for a whole year.
It is a privilege to pursue a PhD within the environment of the Distributed Algorithms CDT. With this project I work at the intersection of academia and industry. As a strong believer in “learning by doing”, this is a unique opportunity to apply my competences as an aerospace engineer while developing expertise on the topic of Machine Learning. It provides me with the prospect of requalifying into a field I would not have accessed by solely pursuing my career in the aerospace industry.
Carole Liao
Back to: Centre for Doctoral Training in Distributed Algorithms