Lucknow: Researchers at Babasaheb Bhimrao Ambedkar University (BBAU) have developed an artificial intelligence-based method to identify oil content in algae, potentially reducing the cost of biofuel production and offering an alternative to conventional food-based feedstocks, reported News 18.
The research, published in Springer Nature’s journal Bioprocess and Biosystems Engineering, was conducted by Dr Ranjan Singh, Dr Soumya Singh and Professor Sunil Singh from the university’s Department of Basic Sciences.
The researchers have developed a software-based framework combining mathematical modelling and artificial intelligence to detect oil deposits in microorganisms through image scanning. Traditionally, identifying and measuring oil content in algae has involved expensive and time-consuming procedures. The new system aims to simplify the process by analysing microscopic images and estimating the oil content.
According to Dr Ranjan Singh, algae could potentially produce 20-30 times more biofuel than conventional crops, making them a promising feedstock for the biofuel industry.
Algae offer several advantages over traditional agricultural feedstocks. Since they are not used as food, their cultivation for biofuel does not directly compete with food production. They can also grow on barren land and in wastewater or saline water, reducing dependence on fertile agricultural land and freshwater resources.
Several algae species can double their biomass in less than 24 hours, while some contain 70-80% oil by dry weight. The extracted oil can be used to produce biodiesel and other biofuels.
The researchers noted that replacing conventional petrol and diesel with algae-based biofuels could potentially reduce greenhouse gas emissions by 40-86%, depending on the feedstock and production process.
The AI-based system requires users to upload a microscopic image of the sample, after which the software identifies oil deposits and estimates their quantity. In a study involving an oil-accumulating yeast, the technology detected oil droplets with 92% accuracy, demonstrating its potential for applications in biofuel research.
The development could help researchers identify high-oil-producing biological feedstocks more efficiently, while reducing the time and costs involved in screening potential sources for commercial biofuel production.
