Fornax - A Unified Smart Campus Management System: Face-Based Attendance, Adaptive Learning, and Explainable Automated Grading

Project name: Fornax -Unified Smart Campus Management System Overview Colleges usually keep attendance, learning progress, and exam results in separate tools. Because of this, it is hard to see whether a student’s absence is connected to weak topics. This project puts all three in one system with one student record. Problem it solves Proxy attendance: Roll calls waste lecture time, and card swipes can be cheated by friends. Same pace for everyone: Students have different knowledge levels, but the course treats them the same. Slow and uneven grading: Teachers spend a lot of time marking descriptive answers, and marks differ between teachers. Main modules 1. Smart attendance (two modes) Photo mode: The teacher takes a classroom photo. The system matches the faces with enrolled students. No student device is needed. Self-marking mode: The student enters a rotating 6-digit code. The system then checks a blink (liveness) and the face, and also checks the college network. Doubtful cases go to the teacher for review. 2. Adaptive learning It tracks a mastery score for each topic using a simple moving average. It recommends study resources, starting from weak prerequisite topics, and it gives a reason for every suggestion. 3. Explainable automated grading It grades typed and handwritten answers using an AI model and a rubric made by the teacher. It shows marks for each criterion, with the proof quoted from the student’s answer. Safety rules: it blocks prompt injection, checks the AI’s output in code, and sends uncertain or handwritten answers to a teacher. The teacher can always override the marks. 4. Risk indicator and teacher dashboard It combines attendance and mastery into a risk level (OK, WATCH, AT_RISK). A heatmap shows which students are weak in which topics. Topics that are not tested are shown as grey (missing), not as low scores. Key features One database linking attendance, mastery, and grades Privacy-friendly: face processing runs in the browser, and only a numeric template is stored, not the photo Human in control: AI is an assistant, and a teacher approves uncertain grades Every teacher change is saved in an audit log Tech stack React, TypeScript, Vite, face-api.js, Node.js, Express, PostgreSQL, Prisma, JWT, and a vision-capable LLM. Current status The system is built as a prototype, and the formulas are checked with worked examples. The experiments with real users are planned as future work.

Authors

Institutions

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-16
DOI
https://doi.org/10.5281/zenodo.24272481
Primary Topic
Artificial Intelligence in Education
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Fornax - A Unified Smart Campus Management System: Face-Based Attendance, Adaptive Learning, and Explainable Automated Grading

Prof. Komal Munde, Karan Danole, Atharv Lugade, Pranshu Jain
2 links
Zenodo (CERN European Organization for Nuclear Research)
Artificial Intelligence in Education
preprint

Fornax - A Unified Smart Campus Management System: Face-Based Attendance, Adaptive Learning, and Explainable Automated Grading

Prof. Komal Munde, Karan Danole, Atharv Lugade, Pranshu Jain
preprint en

Abstract

Project name: Fornax -Unified Smart Campus Management System Overview Colleges usually keep attendance, learning progress, and exam results in separate tools. Because of this, it is hard to see whether a student’s absence is connected to weak topics. This project puts all three in one system with one student record. Problem it solves Proxy attendance: Roll calls waste lecture time, and card swipes can be cheated by friends. Same pace for everyone: Students have different knowledge levels, but the course treats them the same. Slow and uneven grading: Teachers spend a lot of time marking descriptive answers, and marks differ between teachers. Main modules 1. Smart attendance (two modes) Photo mode: The teacher takes a classroom photo. The system matches the faces with enrolled students. No student device is needed. Self-marking mode: The student enters a rotating 6-digit code. The system then checks a blink (liveness) and the face, and also checks the college network. Doubtful cases go to the teacher for review. 2. Adaptive learning It tracks a mastery score for each topic using a simple moving average. It recommends study resources, starting from weak prerequisite topics, and it gives a reason for every suggestion. 3. Explainable automated grading It grades typed and handwritten answers using an AI model and a rubric made by the teacher. It shows marks for each criterion, with the proof quoted from the student’s answer. Safety rules: it blocks prompt injection, checks the AI’s output in code, and sends uncertain or handwritten answers to a teacher. The teacher can always override the marks. 4. Risk indicator and teacher dashboard It combines attendance and mastery into a risk level (OK, WATCH, AT_RISK). A heatmap shows which students are weak in which topics. Topics that are not tested are shown as grey (missing), not as low scores. Key features One database linking attendance, mastery, and grades Privacy-friendly: face processing runs in the browser, and only a numeric template is stored, not the photo Human in control: AI is an assistant, and a teacher approves uncertain grades Every teacher change is saved in an audit log Tech stack React, TypeScript, Vite, face-api.js, Node.js, Express, PostgreSQL, Prisma, JWT, and a vision-capable LLM. Current status The system is built as a prototype, and the formulas are checked with worked examples. The experiments with real users are planned as future work.

Zenodo (CERN European Organization for Nuclear Research)
MIT Art, Design and Technology University (IN)
Quality education
Artificial Intelligence in Education
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon