This paper proposes Quantum Neural Networks (QNNs) as a data-driven approach for predicting fuel consumption. We utilize various layer architecture designs available in the Torchquantum framework, including both entangled and non-entangled circuit designs. In general, QNNs can achieve comparable Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE) with significantly fewer trainable parameters. Neither pure QNNs nor hybrid QNN models exhibit the underfitting tendencies seen in classical neural networks (CNNs). Notably, one of the most significant findings of this work is that hybridizing or”dressing” the quantum circuit leads to substantial improvements in RMSE and MAPE for pure QNNs. These promising results suggest potential optimizations for reducing emissions in green shipping.

This paper proposes a data-driven approach to reduce emissions in international shipping, aligning with the IMO's goal of achieving net-zero greenhouse gas emissions by around 2050. Digital twins (DTs) offer promise for maritime decarbonization due to their simulation and big data handling capabilities. However, fully realizing DTs for new-build is by definition challenging as it requires a real-time data connection. Thus, the research begins with retrofitting existing ships using operational data collected through Bunker Delivery Notes (BDNs), a mandatory method for larger ships since January 2019. The proposed framework constructs digital models to support the retrofit DT, that are tested on a 300m bulk carrier. A fuel consumption model is built using a gray box approach, while various wind-assisted ship propulsion systems are modeled using a white box approach. The study evaluates the design implications and emissions reduction potential of implementing these systems.