Discuss how artificial intelligence is being used to unlock India's historical scientific archives. Illustrate with reference to the Kodaikanal Solar Observatory study (2026).

India holds vast pre-digital scientific records — hand-drawn charts, logbooks and photographic plates — that were long unusable at scale because manual extraction was slow and inconsistent. Artificial intelligence, particularly deep-learning image segmentation, is now converting this dormant heritage into machine-readable datasets, best illustrated by the 2026 Kodaikanal Solar Observatory (KoSO) study [1].

AI as an enabler of archival science

• Digitisation first, intelligence next: KoSO, a field station of the Indian Institute of Astrophysics under the Department of Science & Technology, has digitised over 1.2 lakh solar images, creating the base repository AI could act upon [2][3].
• Cross-domain model transfer: the study applied the U-Net architecture — originally built for biomedical image segmentation — in two steps, detecting the solar disk and then identifying plages (bright magnetically active patches) from suncharts spanning 1904–2022 [1].
• Consistency across heterogeneous records: AI standardised uneven, observer-dependent drawings into a uniform series covering nine solar cycles (15–23), 1916–2007, a task traditional methods struggle with [1].

Scientific and strategic value unlocked

• Validated output: AI-derived plage areas matched KoSO's Ca II K full-disk observations, confirming reliability and allowing gaps in long-term solar records to be filled [1].
• Space-weather preparedness: century-scale activity data strengthens prediction of flares and geomagnetic storms that disrupt satellites, navigation and power grids [2].
• Global standing: the study, led by ARIES with IIA, IIST and Southwest Research Institute, gives India a rare 100-year dataset of value to world heliophysics [1].

Wider replicability

• The same template suits meteorological, oceanographic, census and survey archives, converting colonial-era records into evidence for climate attribution and policy [3].

The KoSO study shows that AI's real contribution lies not merely in new observation but in revaluing what India already possesses. A national programme to digitise and machine-annotate institutional archives, backed by sustained DST funding and open data access, would multiply returns on a century of patient observation — advancing the constitutional duty under Article 51A(h) to develop scientific temper.