Hyderabad Students Develop Sajag AI Child-Safety Tool for 18 Daycare Centres
Two school students have developed an artificial intelligence-based child-safety system that is being deployed across 18 daycare centres in Hyderabad.
The technology, called Sajag AI, analyses CCTV footage to identify potential child-safety incidents and sends alerts to daycare coordinators and monitoring teams.
The project was developed by Class X students Shloak Bihani and Aryaman Taparia.
The initiative highlights how young students are using artificial intelligence to address practical problems in society.
What is Sajag AI?
Sajag AI is a CCTV-based artificial intelligence system designed to help daycare staff identify potential safety incidents.
Daycare centres often have multiple CCTV cameras operating simultaneously. Monitoring every camera continuously can be difficult for staff.
Sajag AI aims to support this process by analysing camera footage and identifying situations that may require attention.
When the system detects a suspected incident, it sends an alert with a snapshot and basic information to the centre coordinators and monitoring team through SMS and email.
How does the AI child-safety system work?
The system connects to existing CCTV cameras through a low-cost computing device.
According to the published report, the current setup uses a Raspberry Pi 5.
The process works in several stages:
- Existing CCTV cameras capture video footage.
- Sajag AI analyses the footage at the daycare centre.
- The system identifies patterns that may indicate a safety incident.
- It generates an alert with a relevant snapshot.
- The alert is sent to the monitoring team through SMS and email.
- Staff can review the alert and take appropriate action.
The system processes video at the centre. The report says that only a snapshot and basic information about a suspected incident are sent to the monitoring team.
How accurate is Sajag AI?
According to COWE India, Sajag AI identified 97.3% of safety incidents in an independent test set, with an overall accuracy of 91.7%.
During early live testing, the system detected eight out of ten staged incidents.
However, the initial testing also generated approximately one false alert every 30 minutes of normal play.
This is an important limitation. An AI alert does not automatically confirm that a child-safety incident has occurred.
COWE has identified reducing false alerts as a key challenge before expanding the system.
The technology still requires longer testing under normal daycare conditions to establish its reliability.
Where is Sajag AI being deployed?
The system is being deployed across 18 Aveksha daycare centres operated by the Confederation of Women Entrepreneurs of India (COWE India).
These facilities support children of women industrial workers and families from underserved communities.
Each centre accommodates approximately 20 to 30 children between six months and six years of age.
The centres provide free or highly subsidised childcare facilities.
The initiative aims to help staff monitor multiple CCTV feeds more effectively while maintaining human supervision.
Who developed the technology?
The project was developed by Shloak Bihani, a Class X student from Hyderabad, and Aryaman Taparia, a student from Noida.
The project was announced during an event in Hyderabad connected with COWE India’s collaboration with the United Nations Development Programme on workplace-based childcare.
The students received recognition for developing the AI-based child-safety solution.
Can AI replace daycare staff?
No. AI-based monitoring should be treated as an additional safety tool, not a replacement for caregivers or trained staff.
Human supervision remains essential because AI systems can miss incidents or generate incorrect alerts.
The technology can help staff identify situations that need attention, but people must review the alerts and make the final decisions.
What does this innovation mean for India’s AI ecosystem?
Sajag AI demonstrates how artificial intelligence can be applied to real-world challenges beyond chatbots and content generation.
Young developers are increasingly exploring AI applications in education, healthcare, safety, agriculture and accessibility.
The project also highlights the importance of building affordable technology that can work with existing infrastructure.
Frequently Asked Questions
1. What is Sajag AI?
Sajag AI is an artificial intelligence-based CCTV monitoring system designed to identify potential child-safety incidents in daycare centres.
2. Who developed Sajag AI?
It was developed by Class X students Shloak Bihani and Aryaman Taparia.
3. Where is Sajag AI being deployed?
The system is being deployed across 18 Aveksha daycare centres operated by COWE India.
4. How does Sajag AI detect incidents?
It analyses CCTV footage and sends alerts with snapshots to centre coordinators and monitoring teams.
5. What is the reported accuracy of Sajag AI?
COWE reported 97.3% incident identification in an independent test set and overall accuracy of 91.7%.
6. Can Sajag AI completely replace human monitoring?
No. It is intended to support staff. Human supervision and review remain essential.
Conclusion
Sajag AI is an interesting example of how school students can use artificial intelligence to address real-world safety challenges.
Its deployment across 18 Hyderabad daycare centres is a promising development, although longer testing is still needed to establish reliability and reduce false alerts.
