Surat student wins prize for tool that predicts groundwater

A school student from Surat has developed a machine-learning project that estimates groundwater availability using historical borewell records, rainfall patterns and soil-related data. The innovation won the national title at the fourth edition of a nationwide technology and innovation challenge for school students, with nine other projects also recognised for addressing problems ranging from water conservation to cybersecurity and assistive technology.

Dheyay Shah of RMG Maheshwari English School created JALAAKAR, short for The Shape of Water: Predict, Alert, and Act. The project combines historical groundwater information with rainfall and soil-related data to generate estimates at the taluka level, with the aim of supporting better-informed water-management decisions.

Shah was declared the national winner at the competition's finale, held at Bharat Mandapam in New Delhi on October 28, 2026. The fourth edition received more than 6,100 student ideas from over 12,000 schools across 705 districts.

How JALAAKAR uses machine learning to estimate groundwater availability

JALAAKAR brings together historical borewell records, rainfall information and soil-related data to assess groundwater availability. It applies machine-learning techniques to identify patterns across these inputs and generate estimates for different talukas.

The project focuses on groundwater assessment, particularly in communities that depend on borewells for their water supply. Its estimates are intended to provide information that could help guide local water-management decisions.

However, the organisers' announcement does not specify the model's prediction accuracy, whether its results have undergone independent validation or whether the system has been deployed in operational water-management programmes. It is therefore best understood as a student-developed estimation tool rather than an established groundwater monitoring system.

Top 10 student projects recognised in the national competition

The national finale recognised 10 projects developed by 13 students through eight individual and two group entries. The innovations addressed groundwater conservation, water purification, women's online safety, hazardous-environment protection, road safety and assistive technology.

The top 10 projects and their developers were:

Rank 1: JALAAKAR: The Shape of Water: Predict, Alert, and Act

Developed by Dheyay Shah of RMG Maheshwari English School, Surat, Gujarat, the project combines historical borewell records, rainfall and soil-related data with machine-learning techniques to estimate groundwater availability at the taluka level.

Rank 2: JalRakshak: Predicting Borewell Failure Before It Happens

Developed by Vidyut Sundar and Shreyan Limaye of Dhirubhai Ambani International School, Mumbai, Maharashtra, the project uses borewell-specific time-series analysis to predict water-level decline by accounting for rainfall recharge and groundwater extraction.

Rank 3: HydroPlastiNet

Developed by Anish Srinivas, Achyutha Pranav V Subramanian and Akhilesh Ram R of Padma Seshadri Bala Bhavan Senior Secondary School, Chennai, Tamil Nadu, the project combines mechanical filtration with bentonite and activated charcoal to trap microplastics and adsorb dissolved pollutants.

Rank 4: Project Unmuted: The Voice of Women

Developed by Anushka Bhattacharya of SNBP International School, Rahatani, Maharashtra, the platform uses AI, cybersecurity and secure data management to provide structured guidance to women facing cybercrime. It also draws on existing deepfake detection technology.

Rank 5: Project IRIS: Infrared Reflective Interface System

Developed by Ayushman Choudhary of Bhavan's B.P. Vidya Mandir, Ashti, Nagpur, Maharashtra, the project combines infrared reflectometry and inertial measurement to track eye direction and head movement, with potential applications in assistive technology.

Rank 6: AVRODH: Automated Verification for Restricted Operator Descent into Hazard

Developed by Shaurya Dubey of Army Public School, Bareilly Cantt, Uttar Pradesh, the project uses safety interlocks, gas-threshold sensing, RFID-based presence detection and authorisation checks to help prevent unsafe entry into hazardous environments.

Rank 7: Smart Device for Instant Helping during Vehicle Accidents

Developed by Abhinav Srivastava of SJS Shaktingar Sonbhadra, Uttar Pradesh, the device uses sensors to detect potential vehicle accidents. It can switch to a backup power source before initiating a voice call to the emergency number 112.

Rank 8: Project DrishtiBodh: LiDAR-Based Smart Safety Collar for Visually Impaired

Developed by Jitesh Ladda of Scottish Academy Parbhani, Maharashtra, the project uses LiDAR sensors to detect nearby obstacles and a microcontroller to trigger directional vibration alerts for visually impaired users.

Rank 9: AquaSentinel AI: Smart Fluoride Safety Device for School

Developed by Sai Adhrit Singh of Dr. Virendra Swaroop Education Centre, Kanpur, Uttar Pradesh, the device uses a colour-based chemical test, light sensing and an offline machine-learning model to estimate fluoride concentration in water.

Rank 10: Neuradrive: Movement for All

Developed by Advik Jain of The Emerald Heights International School, Indore, Madhya Pradesh, the project explores the use of electrooculography and electroencephalography signals to translate eye movements and brain activity into commands for assistive applications.

Student innovation challenge receives ideas from 705 districts

The fourth edition recorded more than 38,000 registrations from over 12,000 schools across 4,500 cities and 705 districts. Participation covered all 28 states and eight Union Territories, including 92 Aspirational Districts.

A jury comprising academicians and industry experts evaluated the top 10 projects on creativity, scientific application, feasibility, cost-effectiveness and potential social impact. The programme offers educational scholarships and rewards worth up to Rs 35 lakh, subject to its terms and conditions.

Alongside the student awards, 10 educators were recognised through the National Teacher Innovation Fellowship. The fellowship received more than 200 applications from 14 states and 26 districts, recognising approaches aimed at making science and technology education more practical and engaging.