A three-woman team at the Birla Institute of Technology (BIT) Mesra in Ranchi has trained AI models to detect and analyse lunar craters, reporting work funded by the Indian Space Research Organisation (ISRO). According to India Today, the project began in February 2023 after a Chandrayaan-2 data analysis workshop and focused on deep learning applied to Digital Elevation Models (DEM). Dr Sanchita Paul said, "Our main mission was deep learning lunar crater detection," and India Today reports the work could support crater dating, navigation planning and future Moon landing missions. India Today describes this as the first collaboration between BIT Mesra computer science and ISRO at the institute.
What happened
India Today reports that a three-woman research team at the Birla Institute of Technology (BIT) Mesra in Ranchi, led by Dr Sanchita Paul with collaborators Dr Mili Ghosh and Mimansa Sinha, has spent three years training AI to detect and analyse lunar craters. The article states the project is funded by the Indian Space Research Organisation (ISRO) and began in February 2023, with roots in a Chandrayaan-2 data analysis workshop held a year earlier. The piece quotes Dr Paul: "Our main mission was deep learning lunar crater detection." India Today reports the team worked with satellite-derived Digital Elevation Models (DEM) and frames the results as potentially useful for crater dating, navigation planning and future Moon landing missions.
Technical details
Per India Today, the team adapted deep learning methods to work with DEM rather than standard PNG/JPEG imagery, and engaged domain knowledge from remote sensing specialists. The article does not provide model architecture, training dataset size, or performance metrics in published form. India Today also notes that the effort responded to an ISRO challenge to automate crater identification, historically performed manually by geologists.
Industry context
Editorial analysis: Automated feature detection for planetary surfaces is an active applied-research area where improvements in accuracy and robustness directly affect mission planning, autonomy, and hazard avoidance. Companies and research groups working on similar problems often need to handle domain shifts between orbital sensors, fuse elevation and optical channels, and validate models against curated geological catalogs and mission telemetry.
What to watch
Editorial analysis: Observers should look for a peer-reviewed paper, public dataset or code release, quantitative validation versus established crater catalogs, and any ISRO technical report that documents methodology and error rates. Adoption signals would include reuse of the dataset or models by other planetary science groups, or citation by mission design teams evaluating landing-site safety and navigation assistance.
Scoring Rationale #
This is a notable applied-AI result for planetary science that demonstrates automation of a previously manual task, but it lacks published metrics and broader validation, limiting immediate impact for practitioners.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.