Decolonizing Artificial Intelligence in Public Health: Co-Designing Health Protection Tools for Male and Female Sanitation Workers in Bangladesh to Advance SDGs 3, 6, and 8

Abstract

Sanitation workers in the Dhaka South City Corporation (DSCC) encounter ongoing occupational hazards, experience frequent injuries, and have limited access to emergency assistance. Their health protection is seldom included in Bangladesh's digital health agenda. This study conducted a mixed-methods study (June–December 2025) to examine whether a designed wearable device, specifically a smartwatch-based safety system, can enhance occupational health and safety through real-time alerts and secure communication with authorities. It is found that quantitative findings revealed high health risks (72%) as well as workplace accidents (65%). It also revealed that a serious communication gap, with 80% of workers lacking access to emergency communication. Our findings show that the risk of accidents is significantly increased if a communication device is absent (OR=3.5; 95% CI: 2.1–5.8). In this study, used qualitative interviews and focus group discussions to show how reporting delays, weak enforcement, and infrastructure constraints routinely turn hazards into major incidents. On the other hand, the study reveals a strong openness (90%) to wearable support. The study, developed under a decolonizing artificial intelligence (AI) approach, considers workers as partners in the design. They prioritize trust, affordability, and locally effective governance over “smart” solutions that are one-size-fits-all and compare implementation barriers in Bangladesh with the experiences of high-adoption locations (e.g., the United States and Canada). The study identifies and explores potential pathways for equitable, scalable deployment and advances SDGs 3, 6, and 8. The study’s findings, guide policymakers and city authorities towards safe, inclusive programs.