With the fast progress in forming more complex electromagnetic (EM) structures with many design parameters and large demand for real-time solution to complex EM problems in embedded devices, the need for a new EM solving approach that can keep pace with the computational requirements has become more imminent. This project aims at developing a novel computationally efficient EM solver which is implementable on systems with limited resources using physics-informed sparse deep neural network that solves partial differential forms of Maxwell's equations without relying on other computational EM solver solutions. The successful candidate will specifically develop signal processing and machine learning algorithms for a real-time electromagnetic solver. The selected candidate will be working within UQ's Electromagnetic Innovations team, led by Professor Amin Abbosh. The candidate will have access to the required simulation tools, and suitable computational resources.