This fully-funded PhD studentship aims to look at what advancements in autonomous vehicles and artificial intelligence can bring to this domain with their teams/swarms. Read more Read less
In the maritime and air domains, there are a number of ways of seducing or effecting incoming weapons with both soft and hard kill methods. Decoys are of a soft kill method which in the main uses an effector to mimic the host platform's signature, be this RF, heat or other. Currently, these decoys tend to be one shot one use and are expended after use.
This research aims to look at what advancements in autonomous vehicles and artificial intelligence can bring to this domain with their teams/swarms. Specifically, one of the envisioned benefits is usage such autonomous systems to mimic host platforms for both defence and structured attack missions including suppression of enemy air defence (SEAD) and dual-role as hard kill end-effectors. As such, teams/swarms of decoys can deploy complex defence and attack patterns rather this be through centralized, distributed, or emergent/decentralized coordination. In that respect, there exists considerable research on the actual coordination architectures, target-task assignment, emergent behaviour, planning, and control aspects of autonomous vehicle fleets and swarms. However, key challenges associated with (a) the spatial and temporal feasibility of mimicking (individually or as a group) actual signature (i.e. IR, RF, or a combination) patterns or cluttering/deception in face of aggressors with a wide range of realistic seeker/sensor and guidance models, (b) deployment of the most effective decoy tactics dependent on the threat type, and (c) ability to organize in real-time based on evolving multi-threat environment and counter-measures are still not fully addressed. In that respect.
AI/Machine learning methods are employed to develop complex surrogate models/digital twins of the threats and the host platform signatures to mimic (depending on the threat/target location and orientation). Subject to any scenario data, characteristics, and performance being deemed too classified for academic research. In addition, utilizing reinforcement learning and its intrinsic game-theoretic basis, complex defence and attack tactics can be learned across a wide range of settings via training driven by a massive number of realistic fast-time simulations of the attack and defence scenarios. Within this context, this study aims to look at the below proposed areas of research:
- A review of existing research work on decoy systems, CONOPS, attack-defence strategies, game-theoretic formulations, sensor/seeker models, IR/RF effectors for mimicking, jamming/spoofing, aggressor models, and guidance strategies.
- Decoy team/swarm deployment and command/control/communication architectures and the effect/sensitivity of such limitations on engagement in dynamic multi-threat/target scenarios both from the defence and also the attack side.
- Using synthetic data to train AI defence and attack models; integration of computational HW and development of models within a fast-time synthetic environment.
- Application of decision tree-based search with reinforcement learning models to achieve generalisation with limited knowledge of the underlying dynamics and environment of the aggresso.
- Creation and parametrized learning/optimization of pre-structured approximate attack and defence tactics.
This research is expected to include theoretical analysis, modelling and computational implementation in association with BAE Systems. It is presumed that the computational demonstration will involve mid-to-high level fidelity models in a realistic synthetic environment which embeds computer generated forces simulation.
Cranfield is an exclusively postgraduate university that is a global leader for education and transformational research in technology and management. This PhD will be hosted by the Centre for Autonomous and Cyber-Physical Systems. The Centre for Autonomous and Cyber-Physical Systems is one of the world’s largest centres of postgraduate education and research, with over 200 MSc and PhD students.
You will be encouraged and supported in publishing your own work in high-quality peer-reviewed journals. Also, you will have opportunities and supports to present your work at the relevant UK and international conferences. You will obtain knowledge on the technologies for the related disciplines, experience the procedures of algorithm development in autonomy and AI, and learn skills for modelling, synthetics, and simulations.
At a glance
- Application deadline15 Oct 2021
- Award type(s)PhD
- Start date10 Jan 2022
- Duration of award4 years
- Reference numberSATM232
Applicants must have a B.Sc. and a Master’s degree, both in engineering or a related area. A demonstrated background in aerospace, autonomy, and AI/ML would be a distinct advantage.
This is a fully-funded opportunity.
Sponsored by BAE Systems through ICASE 2021, this studentship will provide a full bursary for 4 years up to £18,000 per annum depending on qualifications, plus full tuition fee. The application is open to UK citizens and UK residents.
About the sponsor
Sponsored by EPSRC and BAE Systems, this studentship will provide a bursary of up to £18,000 (tax free) plus fees* for four years.
Cranfield Doctoral Network
Research students at Cranfield benefit from being part of a dynamic, focused and professional study environment and all become valued members of the Cranfield Doctoral Network. This network brings together both research students and staff, providing a platform for our researchers to share ideas and collaborate in a multi-disciplinary environment. It aims to encourage an effective and vibrant research culture, founded upon the diversity of activities and knowledge. A tailored programme of seminars and events, alongside our Doctoral Researchers Core Development programme (transferable skills training), provide those studying a research degree with a wealth of social and networking opportunities.