General information
Title: Sustainability-focused agent-based modelling for aircraft design with manufacturing flexibility
Supervisors: Dr Thulasi Mylvganam, Professor Rob Hewson, Dr Ajit Panesar, Dr Rhea Liem
Reference code: BI0002
Project overview
Aircraft design exists as a component of a hugely complex and dynamic air transport system. Uncertainty and difficult-to-quantify factors make the process highly challenging. Real demand is stochastic, with rapid ramp-ups, controlled ramp-downs, single-source outages and lead-time spikes all increasing system complexity and uncertainty. Current design practice treats manufacturing flexibility as an afterthought rather than an optimisation variable. The result is costly redesign, delivery slippage and aircraft-on-ground risk.
This project seeks to redefine how large-scale engineered systems are designed, not just to perform, but to be built adaptably and sustainably. The research will formulate both the design environment and the surrounding factors that influence it as a multi-agent system, enabling the impact of design and manufacturing decisions to be assessed, quantified and ultimately designed for.
By combining optimisation, multi-agent simulation and machine learning, the project will develop methods that deliver tangible and measurable benefits for manufacturers and airlines through reduced ramp-up delays, aircraft-on-ground risk and redesign cost.
Research objectives
The project will:
- Formulate aircraft design and manufacturing environments as a multi-agent system in which the impacts of design decisions can be assessed, quantified and incorporated into the optimisation process.
- Develop a sustainability-embodied agent-based framework with agents representing subsystems (spars, ribs and skins), resources (stations, jigs and robots) and suppliers (multi-source parts, materials and processes).
- Model interactions and negotiation between agents across performance, schedule and supply-chain objectives.
- Develop manufacturing-aware design parameterisation and optimisation methods within a dynamically reconfigurable production environment.
- Define suitable measures of merit that support sustainable aviation goals while balancing performance, manufacturability and resilience.
- Investigate computational approaches, including multi-fidelity modelling and adaptive learning, to make the decision-making framework computationally tractable.
Contact
We welcome enquiries from prospective applicants interested in the project or the doctoral programme. Please use the contacts below, depending on the nature of your enquiry.
Doctoral programme related enquires
brahmal-aviation@imperial.ac.uk
Project related enquires
- Dr Thulasi Mylvaganam (primary supervisor)
- Professor Rob Hewson
- Dr Ajit Panesar
- Dr Rhea Liem
