Journal of Engineering Research and Reports

The utilization of inedible agricultural seeds to make bio-oil is not new, but accelerating its usefulness via potential options is ideal. Therefore, studies showing that inedible seeds are superior to edible seeds were examined from 2011 to 2023. For additional research, an ideal inedible seeds known as yellow oleander seeds were selected. From the review: ANN performed better in the best extraction of bio-oil than the D-Optimal Design of RSM. This is because artificial neural networks (ANNs) have a great deal of strength when it comes to modeling the production of biofuel and bio-oil because of their diverse network topology, quick learning algorithm, high error tolerance for non-linear processes, and flexible learning method. Nonetheless, both are effective modeling and optimization instruments for extracting oil from yellow oleander seeds.  Furthermore, it has been established that the Soxhlet extraction method outperforms non-conventional approaches like enzyme-assisted extraction, supercritical fluid extraction, ultrasound, and microwaves, as well as the traditional approach of mechanical extraction. These novel non-conventional methods can increase oil extraction rates, decrease oil quality degradation, and shorten extraction timeframes. Additionally, they have been effectively employed to mitigate certain disadvantages linked to traditional techniques for extracting oil; yet, Soxhlet extraction continues to produce the highest output. Since the use of other machine learning models apart from ANN has not yet been documented, further research is required to determine whether these programs can be used to model and optimize the extraction of yellow oleander oil. Hence, this will help the optimizing industry to generate more income for the Country. Note, the limitations of this study boils down to how to utilize other machine learning models apart from ANN for optimization and the adaptability of the bio-oil seeds in frozen regions.