Jakarta, INTI - Grab Indonesia has developed artificial intelligence (AI) and machine learning–based technology to monitor the implementation of the Free Nutritious Meal Program (MBG) in real time. The MBG initiative run by Grab is part of the company’s corporate social responsibility (CSR) program and is fully funded by the private sector.
Grab Indonesia’s Head of Safety & Quality, Sherylin, explained that the system is designed to detect potential risks early, before meals reach students.
The adoption of technology, she added, was driven by the limitations of manual monitoring. With AI-powered systems, supervision can be conducted faster and with greater accuracy.
“Why technology? Because we see that manual monitoring has gaps. With technology, we can monitor in real time, and we can mitigate risks before problems reach our younger brothers and sisters at school,” Sherylin said during a visit to the MBG Command Center at Poins Mall, South Jakarta, on Tuesday, February 10, 2026.
Smart CCTV for Anomaly Detection in Partner Kitchens
The monitoring system operates through CCTV cameras installed in three critical areas of partner MSME kitchens: the cooking area, the serving area, and the handover point with delivery drivers.
According to Sherylin, the cameras do more than simply record footage, they are equipped with AI capabilities that automatically detect various anomalies.
“The CCTV is not just recording anymore, but has been enhanced with AI technology, allowing us to detect anomalies. We developed this technology ourselves,” she said.
One of the key features is the detection of personal protective equipment (PPE) used by food handlers. The system can identify whether kitchen staff are wearing gloves, masks, hairnets, and aprons in accordance with safety standards.
“If kitchen staff are not using them properly, we immediately receive alerts. There are several visual boxes that appear, which is also a technology we developed to help monitoring agents see what objects are being detected, along with a confidence level showing how certain the system is about its detection,” she explained.
In addition, the system is capable of detecting pest activity such as rats, cockroaches, and geckos in kitchen areas. When detected, alerts are sent so teams can quickly intervene.
Sherylin added that the technology is developed internally and continuously trained to improve its accuracy. She acknowledged that inaccuracies occurred in the early pilot phase, particularly due to low confidence levels in detection.
“Has it ever been inaccurate? Of course, especially in the early pilot stage. But we keep improving every day by training more correct and incorrect examples so the machine learning becomes much smarter,” she said.
Conclusion
Grab’s AI-based monitoring system demonstrates how artificial intelligence can strengthen food safety and operational transparency in large-scale social programs. By combining real-time surveillance, automated anomaly detection, and continuous machine learning improvements, the initiative highlights the growing role of AI in ensuring quality, efficiency, and public trust in community nutrition programs.
Read more: ChatGPT Introduces Advertisements