evidoria

← Back to browse

Good practice Imported

Nirantara — Karnataka's Statewide AI Facial-Recognition Attendance Rollout for 5.2 Million Students

India · Bengaluru · See the India profile · See the Bengaluru profile

Evidence: Descriptive / self-reported Top 82% 33/100 · Ask Evidence Copilot about this practice

Karnataka's education department is rolling out 'Nirantara,' an Aadhaar-linked facial-recognition attendance app, to 52,686 government/aided schools (5.2M+ students) after a 2025 pilot in Uttara Kannada. No independent accuracy or efficacy evaluation has yet been published.

52,686 schools
Schools covered by phased rollout
46,460 schools
Government schools covered
6,226 schools
Aided schools covered
5.2 million+ students
Students covered by rollout
50 students
Students' attendance registered per single capture (within seconds)
Nirantara — Karnataka's Statewide AI Facial-Recognition Attendance Rollout for 5.2 Million Students

Details

Maturity
Scaling
Promoter
Department of School Education and Literacy, Government of Karnataka
Period
2025–2027 (Uttara Kannada pilot 2025-26; statewide rollout 2026-27)
Keywords
biometric attendance, facial recognition, school administration, welfare-scheme tracking, education data systems

Context

In its 2025-26 state budget, Karnataka's Chief Minister Siddaramaiah announced 'Nirantara' ('continuous'), an AI-based facial-recognition attendance system for the state's Department of School Education and Literacy, intended to replace manual roll calls across all government and aided schools. The system converts each student's photo into an encrypted, vector-based biometric ID, integrated with the state's existing Students Achievement Tracking System (SATS) and linked to Aadhaar, India's national identity system. Software and data are hosted at the Karnataka State Data Centre (Vikasa Soudha, Bengaluru).

Objectives

The primary goal is to automate daily attendance capture in place of manual roll calls. Officials also plan to use the same mobile app to track delivery of welfare schemes such as midday meals and free footwear.

Activities

Teachers use a dedicated mobile app to photograph the classroom, which the department says can register the attendance of 50 students within seconds, including in low-light conditions. Following a pilot in Uttara Kannada district that officials described as successful, the state approved a phased rollout beginning in the 2025-26 academic year.

Results

The approved rollout covers 52,686 schools — 46,460 government and 6,226 aided institutions — reaching more than 5.2 million students, with full statewide coverage targeted for 2026-27. As of mid-2026, no independent third-party study of recognition accuracy, error rate, or effect on attendance or learning has been published.

Conclusions

This contrasts with comparable systems elsewhere, such as Paraná, Brazil, where an academic accuracy audit and a teachers' union survey were later commissioned. Officials describe the encrypted-ID design as a privacy safeguard, but no data-protection impact assessment has been made public, and collecting biometric data from more than 5 million children — tied to Aadhaar — raises unresolved consent and retention questions. At 5.2 million students, Nirantara would be among the largest deployments of biometric school-attendance technology anywhere.

Implementation

Indicative cost
High (€500k–€5M) — No public budget figure disclosed; infrastructure is state-hosted at the Karnataka State Data Centre.
Time to results
Medium (1–3 years) — Pilot in Uttara Kannada district in 2025-26; phased statewide rollout to 52,686 schools targeted for full coverage by 2026-27.
Staffing & skills
Department of School Education and Literacy, Government of Karnataka, Teachers use a dedicated mobile app to capture classroom photos for attendance

Conditions for success

  • Integration with existing state systems (Students Achievement Tracking System / SATS) and Aadhaar national ID
  • Centralized hosting at the Karnataka State Data Centre (Vikasa Soudha, Bengaluru)
  • A completed district-level pilot (Uttara Kannada) before phased statewide rollout

Common failure modes

  • No independent third-party study of recognition accuracy, error rate, or effect on attendance/learning has been published as of mid-2026
  • No public data-protection impact assessment despite biometric collection from over 5.2 million children linked to Aadhaar
  • Consent and data-retention questions for minors' biometric data remain unaddressed in public reporting
  • Contrasts with comparable systems elsewhere (e.g., Paraná, Brazil) where an academic accuracy audit and a teachers' union survey were later commissioned — no equivalent independent scrutiny yet exists for Nirantara

Where it fits

Governance type
state government (Karnataka Department of School Education and Literacy)
Scale
statewide
Income level
lower-middle-income (India)

Commonly funded by

National / regional programmes

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful