Researchers at the University of Oxford’s Bodleian Library have utilized a new AI-driven multispectral imaging system to decipher 8th-century Sharada script palimpsests found in a Himalayan cache. The hidden layers of these manuscripts reveal a lost Sanskrit treatise dedicated to Vrikshayurveda, specifically focusing on plant pathology. The text details sophisticated protocols for identifying and treating various fungal infections in cereal crops using botanical extracts and mineral compounds.
This breakthrough in manuscript research underscores the depth of ancient Indian botanical science. The AI model, trained on thousands of script variants, successfully reconstructed 85% of the fragmented text, which includes systematic classifications of plant diseases based on moisture and soil quality. Scholars believe these findings could offer "green" alternatives for modern agricultural pest management by revisiting ancient bio-pesticide formulations used in early medieval India.