Estimate and optimize fuel consumption on vessels to reduce CO2 impact of fishery

摘要

Effectiveness of fuel consumption in fishery shipping logistics is needed to minimize costs and CO₂ emissions. This paper proposes a way to combine current internal (sensor data) and external (satellite) observations of fishing vessels to optimize routing between several target locations and consecutively find the optimal speed for each segment. Therefore, a predictive model (estimated fuel consumption) and a gradient descent algorithm (optimal speed) have been developed. A real-world case has been used to describe the potential usage and benefits of data and predictive analysis in the fishery domain. In particular, estimations are calculated, which suggest improvements can be made with regards to fuel consumption. Methodologically, the predictive model is mainly based on polynomial ridge regression.