The HealthBot project was born from a desire to address a critical challenge evident in many Greek island communities, including our own in Kos: the gap in accessible, preventative healthcare.
Motivated by this local need and inspired by a national student competition, our goal was to explore how technology could empower individuals to take a more proactive role in their well-being. This initiative was designed with a dual purpose: to create a tangible tool for health equity and to serve as a powerful, real-world case study for Project-Based Learning (PBL) in STEM.
The Solution: An AI-Powered Health Kiosk
At its core, HealthBot is an autonomous health kiosk that integrates a suite of sensors and artificial intelligence to provide immediate, non-diagnostic health insights. The system is designed to measure key parameters such as weight, height, and advanced biometric data, including heart rate (HR), blood oxygen saturation (SpO2), and respiration rate (RR), using a specialized MAX86150 biosensor.
Furthermore, it leverages AI in two significant ways: a cloud-based machine learning model (ModelDerm) offers preliminary analysis of skin lesions, and a Large Language Model (GPT-4 mini) synthesizes all the collected data to provide personalized, easy-to-understand wellness advice.
The Journey: Student Innovation in Action
This project was, first and foremost, an educational endeavor driven by a talented team of high school students from the Instudies center: Filippos Papanikolaou, Tasos Syris, and Giota Nikoloudaki. Over a six-month period, they drove the entire development process, from conceptual design to a functional prototype. Guided by the principles of PBL, they tackled significant technical hurdles that required remarkable ingenuity. This included not only the mechanical construction and hardware integration but also developing custom software from scratch to interface with the advanced MAX86150 biosensor, as no ready-made libraries were available for their platform.
We experimented with different materials and calibration algorithms to optimize the accuracy of the weight scale, demonstrating a deep, hands-on engagement with the engineering process.
This journey allowed them to cultivate crucial 21st-century skills in problem-solving, collaborative design, and system development.
Validation and Future Vision
The team’s work culminated in a public testing event in Kos, where 24 community volunteers provided valuable initial feedback.
The results from the NASA-TLX workload assessment were promising, indicating the kiosk was generally easy and not burdensome to use. The feedback also highlighted key areas for technical refinement, such as final sensor calibration, providing critical insights for the next development phase.
HealthBot stands as a proof-of-concept demonstrating that student-led innovation, when channeled through effective educational frameworks like PBL, can produce tangible technological solutions for real community needs. In recognition of its dual value in health technology and education, our paper detailing this project was accepted for presentation at the 14th Panhellenic Conference on ICT in Education. The future of this work lies in further technical validation and exploring sustainable pathways for deployment in the communities it was designed to serve.

