Jakarta, Indonesia Sentinel — Four Indonesian students from Binus University Alam Sutera have developed an AI-powered interactive doll named PawPal, designed to serve as a smart, conversational playmate for children. Their motivation stems from growing concerns over toddlers’ increasing screen addiction, particularly the use of smartphones.
The PawPal prototype was unveiled during the Samsung Innovation Campus Batch 6 2024/2025 competition, organized by Samsung Indonesia. The invention took home top honors in the university category of the digital talent development program.
The team behind PawPal, called Rarevolution, consists of four computer science students from Binus. They were selected as the best university team following a rigorous selection process that included multiple learning stages and mentorship sessions with industry and academic experts. The winners were officially announced on Wednesday, May 28, 2025.
Rarevolution focused on Indonesia’s alarming screen time habits. According to 2023 research by Data.ai, Indonesians average 5.7 hours of screen time per day. It was an issue that extends to children and has been linked to negative developmental outcomes.
Studies show prolonged screen exposure is associated with decreased functional connectivity in parts of the brain responsible for language and cognitive control.
Read Also:
Researchers Warns Growing Threat of Toxic Fungi as Climate Crisis Worsens
PawPal aims to offer a screen-free alternative to help children learn and interact through speech. The AI- and IoT-integrated doll features a microphone, mini computer, and speaker.
When a child speaks, the built-in microphone captures the audio, which is transcribed into text, processed by a large language model (LLM), and converted back into speech, enabling real-time, interactive dialogue.
The AI-Powered doll comes equipped with several features, including “Talk to Me,” “Math Adventures,” “Would You Rather,” and “Guess the Sound.”
Currently, PawPal operates using a Malay-language LLM, but team member Rowen Nicholas, who serves as the team’s Machine Learning Engineer, plans to integrate an Indonesian-language model to make responses more natural and locally relevant.
(Raidi/Agung)
