Guwahati: Tea factories have long grappled with the age-old challenge of perfecting the art of withering—a pivotal process determining tea quality.
Withering, the initial step in manufacturing black tea, is crucial for producing high-quality black tea. It begins with spreading freshly harvested tea leaves in a proper withering system to lose moisture before processing.
In many tea factories, withering is gauged subjectively by factory staff based on experience, often resulting in time-efficient but subjective decisions.
A change is brewing as experts advocate for digital interventions to support precision withering in tea factories as part of a larger ecosystem transformation. Some tea estates in Assam have already initiated digital interventions.
Scientists from Tata Consultancy Services Research and Innovation, Mumbai, conducted a study titled “Towards Precision Withering in Tea Factories with Non-invasive Leaf Moisture Estimation.” They utilized spectral data from a sample of leaves to estimate withering levels based on inherent moisture levels.
The study, conducted at Koliabur tea estate in Assam in partnership with Fasttrack Agrotech LLP, utilized value-added spectral cameras.
Spectral data of tea leaves measures light absorption, reflection, or emission across different wavelengths, providing insights into various leaf characteristics, such as chemical composition and health status.
Compared to traditional methods, spectral estimation offers a more efficient and cost-effective solution, enabling tea factories to streamline operations, reduce labor costs, and improve productivity.
Moreover, spectral analysis allows continuous monitoring of moisture levels in tea leaves throughout the withering process, providing real-time feedback to operators for timely adjustments to optimize withering levels and maintain tea quality.
“Withering is crucial in making quality teas. The challenge lies in determining moisture percentage during withering and achieving an even wither. Using spectral imaging and analysis, this process could be made more precise,” a stakeholder in the project told EastMojo.
He added that spectral data analysis through algorithms indicates achieved withering levels.
“The spectral camera measures moisture levels in withered leaves much faster compared to traditional methods like microwave ovens,” an official of Kaliabor tea estate said.
The study defined suitable moisture ranges for Indian tea processing and developed models to predict withering levels, with the Support Vector Machine (SVM) model proving most effective.
“Performance of the models could be further improved by introducing appropriate dataset variations and exploring deeper learning methods. This intervention could significantly benefit digital farming and food processing, particularly in the tea ecosystem,” the study concludes.
Also Read: Meghalaya: KSU members held in murder cases remanded to police